# Mould Detect: full site text > AI mould detection and mould-risk forecasts for Australian homes, and Mycelium, the MCP server and CLI that give AI agents Mould Detect's data. Generated from the published pages; the index is https://moulddetect.ai/llms.txt. --- # About Mould Detect | The People Behind the Instrument URL: https://moulddetect.ai/about/ ABOUT US · ESTABLISHED 2026 ## Environmental science, brought home. Mould Detect helps people find mould early, understand what it means for their health and their home, and get it fixed by the right people. One photo gives an AI reading. From there the app connects you to Mould Edu, advice from Miss Mould, lab testing with approved partners like CDES, and soon, vetted remediation specialists. OUR MISSION ### Holistic solutions to human health challenges, with a focus on improving the environments people live in. It starts with healthier homes. Mould is the new asbestos. MIRKO NAD · CO-FOUNDER WHAT GUIDES US ### Three commitments. A natural core. Remediation that works with nature, not against it. Non-toxic, preventative, and proven. AI with expertise behind it. Advanced detection models built alongside mycological labs and working remediation professionals. Proactive care. Finding issues before they reach your lungs, your walls or your bond. THE FOUNDING TEAM · 7 PEOPLE ### Seven people who know buildings, health and AI, building one place to find, understand and fix mould. Three co-founders bring remedial building, natural remediation and AI platform experience. Four founding specialists bring AI engineering, cloud and IoT, AI quality and growth. #### CO-FOUNDERS - ##### Mirko Nad CO-FOUNDER · CHIEF EXECUTIVE OFFICER A remedial builder for more than twenty years, with over 1,500 remediation projects. Mirko founded Keystone Building Solutions and Riteflo, maker of ventilation systems and Damp Rods, and wrote The Seven Most Common Problems in Strata Buildings. Mirko keeps what the app recommends true to what works on a real job. - ##### Rozario Chivers CO-FOUNDER · CHIEF TECHNOLOGY OFFICER More than twenty years in digital technology and AI, across enterprise SaaS, data, eCommerce and fintech. Rozario co-architected our edge and cloud AI, built the environmental data platform and specialists directory, and co-designed our ultrasound diffuser and IoT sensors. - ##### Norah Copithorne CO-FOUNDER · CHIEF RELATIONSHIP OFFICER Founded Nature's Defence in December 2023 to treat mould with natural, chemical-free products, built on Dr Edward R. Close's research and a proprietary oil range. Norah's field protocol tests before and after every treatment, and a grassroots network of cleaners, remediators and builders are among our first users and data contributors. #### FOUNDING SPECIALISTS - ##### Anna Huang FOUNDING AI ENGINEER The original architect of our AI vision platform and of Miss Mould, the AI mycologist, which uses an LLM as a judge to check its sources before it answers. Master's in AI from UTS, with cloud and edge AI and data governance experience. - ##### Vladimir Lekic CLOUD ARCHITECTURE AND EMBEDDED SYSTEMS More than twenty years of enterprise architecture in insurance, financial services and government. Vladimir designs our cloud platform, data lake and the embedded systems inside our IoT sensors. - ##### Manoj Kumar AI OPS AND QUALITY ENGINEERING More than twenty years in software quality, VP Technology at Moolya and a member of the Selenium Project Leadership Committee. Manoj tests AI that can answer differently each time, so results stay dependable as our models improve. - ##### Ming Johanson MARKETING AND GROWTH Evidence-driven marketing, branding and growth, with martech and AI experience. Director of JOGO Digital, business coach and startup mentor. ### Join Mould Detect. We are looking for health-obsessed product, engineering, user experience and AI people, and for partners who want healthier homes as much as we do. Software Engineering · User Experience Design · AI Research and Development · Human and Building Health [Get in touch](https://moulddetect.aimailto:careers@moulddetect.ai)[Join the early access](https://moulddetect.ai/#early-access) --- # Mycelium: MCP Server and CLI for AI Agents | Mould Detect URL: https://moulddetect.ai/agents/ AI AND AGENT FIRST ## Mould Detect, built for agents . Mycelium Network connects AI assistants and your own agents to Mould Detect's data: the mould growth index for a location, the weather history behind a damp wall, river flood forecasts, air quality, and a mould check for photos. Every answer carries its sources. Mycelium CLI ### Three ways in Miss Mould is Mould Detect's own first client of the same Mycelium tools. In the Miss Mould chat at mycology.moulddetect.ai, when you share your location she gives live local weather and mould risk for your area, with sources. It is decision support from outdoor conditions. #### MCP server One endpoint at mcp.moulddetect.ai for Claude, ChatGPT, Cursor, Codex and any MCP client. Our own Miss Mould uses it too. #### Developer portal Free beta keys, setup for each client, and the tools reference at mycelium.moulddetect.ai. #### Mycelium CLI A terminal agent for macOS, Linux and Windows that runs on a free local model or your own AI key. MCP IN USE ### Ask in plain language Add Mould Detect to Claude Code with one command, then ask a question. Claude picks the right tool, calls it, and answers with the numbers and their limits. Real Claude Code session using the Mould Detect MCP server, 21 October 2026. claude mcp add --scope user --transport http mould-detect https://mcp.moulddetect.ai/mcp --header "Authorization: Bearer $MYCELIUM_KEY" MYCELIUM CLI ### Or use the terminal Direct commands need no AI model. Run mycelium on its own to chat with a free local model through Ollama, or your own Anthropic, OpenAI or Google key. The first run signs you in; type / for commands, and the status line shows your model, context use and today's allowance. $ mycelium ▄ ▄ █ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ █ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ █ ▀ ▀ ▀ ▄ ▄ ▄ ▀ ▀ █ ▄ ▄ ▄ █ ▀ ▄ █ █ █ █ █ ▄ ▀ █ ▄ ▀ ▀ ▀ ▀ ▀ ▀ ▄ █ ▀ ▀ █ █ █ █ █ ▀ █ █ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ █ ▄ █ █ █ ▄ ▀ █ ▄ █ ▄ ▄ ▄ ▄ ▀ █ █ █ ▀ ▀ █ ▀ ▀ ▀ ▀ ▀ ▀ ▄ █ ▀ ▄ █ █ █ █ █ █ ▄ ▄ ▄ ▄ ▀ █ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ █ █ █ █ █ █ █ █ █ █ █ █ █ █ ▄ █ █ ▄ █ █ ▀ ▀ ▀ █ █ ▀ ▀ ▀ ▀ ▀ ▀ █ █ ▀ █ █ █ █ █ █ █ ▄ ▄ ▄ ▄ ▄ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▄ ▄ ▄ ▄ ▄ █ █ ▄ ▄ ▄ ▄ █ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ █ ▄ ▄ ▄ █ ▀ ▀ ▀ ▀ ▀ ▀ ▀ █ ▄ ▄ █ ▀ ▀ █ █ ▄ █ ▀ ▀ ▀ █ █ ▄ ▄ ▄ ▄ ▄ ▄ █ ▀ ▀ █ ▀ ▀ ▄ ▄ ▀ ▀ █ █ ▀ ▀ ▀ █ █ █ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ █ ▄ ▀ ▀ █ ▄ ▄ ▀ ▄ ▄ ▄ █ █ ▀ ▀ ▀ █ ▄ █ ▄ ▀ █ ▀ ▄ █ █ ▄ ▄ █ █ ▀ ▄ █ ▄ ▀ █ ▄ ▄ █ ▄ ▄ ▄ ▀ ▀ █ █ █ █ █ ▄ ▄ ▄ █ █ ▀ ▄ █ █ █ █ █ ▄ ▄ ▄ ▄ ▄ ▄ █ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ █ █ █ ▄ ▄ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ ▀ █ █ █ █ ███ ████ █ ███ █ █ █ █ ███ █ ███ ██ ██ █ █ █ █ █ █ █ █ ██ ██ █ █ █ █ █ █ █ █ ███ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ ███ ████ ████ ███ ███ █ █ ███ ████ ███ powered by moulddetect.ai Mycelium CLI v0.3.4 · free beta ollama/qwen2.5:7b · mcp.moulddetect.ai · key …kuDG › Will mould grow in a Hobart basement this week? ● get_mould_risk_index (-42.88, 147.33) ● The Mould Growth Index for Hobart's basement this week indicates no visible mould growth is expected. Here are the details: • Today's Index: 0 (Dry, no growth) • Index in 7 Days: 0 (Dry, no growth) • Trend: Steady • Peak Over 30 Days: 0 (Dry, no growth) • Estimated Days Until Visible Growth: Not forecasted However, it's important to note that this is decision support only and not a medical, laboratory, or building-inspection result. Outdoor conditions do not fully measure indoor mould risk. Consult a qualified professional if you're concerned about mould in your basement. Sources: Contains modified Copernicus Climate Change Service information 2026 · NOAA Global Forecast System (GFS), public domain Decision support only. Not a medical, laboratory or building-inspection result; outdoor conditions do not measure indoor mould. Consult a qualified professional before acting. ──────────────────────────────────────────────────────────── › Try "Is Bondi damp this week?" ──────────────────────────────────────────────────────────── ollama/qwen2.5:7b │ ctx 14% of 32k │ mcp.moulddetect.ai Today ● ● ● ● ● ● ● ● ● ● 97% left 16 of 500 units /help commands · ↑↓ history · tab completes · ctrl-d quits Real output from v0.3.4 on a free local model, 22 October 2026, shown with the current banner and name. Your numbers will differ. curl -fsSL https://mycelium.moulddetect.ai/install.sh | sh TOOLS ### What agents can call Eight read-only tools. get_mould_risk_index Mould Growth Index 0 to 6, global forecast, history in Australia get_weather_forecast 7-day forecast, global get_current_conditions Now, global get_weather_history ERA5-Land history, Australia get_flood_risk River flood forecast with return periods, Australia get_air_quality Air quality against WHO guidelines, Australia data_platform_status Data freshness, free analyze_mould_image Has mould, no mould or uncertain, 20 units [Full tools reference](https://mycelium.moulddetect.ai/docs/#tools) BETA ### Free while we learn The beta is free: 500 units per account per day (an image analysis costs 20, other tools 1), 60 requests a minute, up to three keys. Results are decision support, not medical, laboratory or building-inspection findings, and every result carries its data credits. [Get a free key](https://mycelium.moulddetect.ai/keys/)[Developer docs](https://mycelium.moulddetect.ai/docs/)[Environmental data API](https://environment.moulddetect.ai/docs/)[llms.txt](https://mycelium.moulddetect.ai/llms.txt) AGENT FIRST ### Everything a person can see here, an agent can read. Three machine-readable ways in, for assistants, coding agents and your own tools. - Beta #### The MCP server. Mould risk index, weather history and forecast, flood risk, air quality, current conditions and image analysis, as read-only tools that cite their sources. [TOOLS REFERENCE ](https://mycelium.moulddetect.ai/docs/) - Live #### llms.txt. A plain-text map of this site and a full-text Markdown edition, so a model can read Mould Detect without rendering a page. [LLMS.TXT ](https://moulddetect.ai/llms.txt) - Live #### The data API underneath. The same environmental and weather data over HTTPS, for agents that call APIs directly. [DEVELOPERS ](https://moulddetect.ai/developers/) Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ### Give your agent the mould data. Create a free beta key, add the server to your client, and ask your first question. [Get a free key](https://mycelium.moulddetect.ai/keys/) --- # AI Vision · Mould Detect URL: https://moulddetect.ai/ai-vision/ AI VISION ## Teaching a machine to read mould. By the end of this page you will know what our vision looks at in a photograph, what it learned from, how it turns a picture into a named mould family with a confidence, and where it stops and asks you for a better photograph. THE HARD PART ### A photograph shows you less than you think. Ask a laboratory what is growing on a wall and the answer comes back in two parts, like a name. Aspergillus is the surname: a family, with family habits, family preferences and a family reputation. The species is the first name, and it is often not decidable from a photograph at all without a microscope. So our vision reports the genus, reports how confident it is, and is honest about where the photograph stops being enough. Colour is the first thing people reach for and the least reliable thing in the frame. One genus grows olive on painted plasterboard, close to black on saturated paper facing, and a flat dust grey once it has dried out. Warm room light shifts it again. So does the phone, which sharpens and colours to taste before you ever see the file. Your eye corrects for all of that without being asked. Software does not. Then there is the harder problem, and it is not about mould at all. Nobody tells a classifier what to look at. It is shown a photograph and the right answer, thousands of times over, and each time it nudges itself to be a little less wrong. There is no stage where a person points into the frame and says: that greenish patch, there, that is the thing. So it hunts for whatever reliably predicts the answer, and it takes whatever is cheapest. Think of a new starter handed a stack of photographs and told which drawer each one belongs in, with no explanation. If every photograph bound for the third drawer happens to have a blue folder in the corner, they will file by blue folder. It works, it is quick, and nobody told them the folder was irrelevant. Software is that new starter, except it never asks a question and never tells you which rule it settled on. Put it in a building and the trap is obvious. If one genus is nearly always photographed on a bathroom ceiling, the classifier has two questions available. What is the shape, texture, edge and colour of this colony, which is hard. Or: is this a ceiling, which is easy. Ceilings are flat, pale, evenly lit and shot from below. If the easy question predicts the answer just as well, the easy question is what gets learned. It is the student who revised from one practice paper where every long answer happened to be C. They scored beautifully, the marking was honest, and the result meant nothing at all. And the confidence is the dangerous part. A model that says it is not sure is a nuisance. A model that is fluently, calmly, articulately wrong will be believed, by a client, by an insurer, possibly by you at four o'clock on a Friday. Everything on the rest of this page exists to close that trap. None of it is clever. It is discipline about what goes in, because nobody gets to tell the model what to look at. The photographs are the instruction manual. HOW IT LEARNS ### One photograph becomes many lessons. Every tile below is the same specimen, altered in a way that does not change the answer. The network is shown all six and learns what survives the alteration. - ORIGINAL - ROTATE 90° - MIRROR - CROP + ZOOM - CONTRAST - HUE SHIFT The same drawn specimen six ways: as photographed, turned, flipped, cropped close, pushed for contrast and shifted in colour. The mould is identical in every tile. Only the photograph changed, and the answer must not. You recognise a friend from the side, in a hat, at dusk, half turned away. You are not matching a stored photograph. You learned the structure of a face and everything else fell away as accident. Augmentation teaches a network the same indifference. One photograph is fed in many times over, turned, flipped, cropped, brightened and colour shifted, with the answer held constant every time. The effect is quietly profound. Rotation says the answer does not live in which way is up. Cropping says it does not live in how much of the frame the colony fills, which matters enormously in practice: a survey shot from the doorway and a macro of the same patch do not look remotely alike to software. Colour and contrast shifts say it does not live in the light bulb. What is left, once all of that has been declared irrelevant, is the structure of the growth itself. None of this is proprietary. It is textbook practice, surveyed thoroughly by [Shorten and Khoshgoftaar in the Journal of Big Data](https://journalofbigdata.springeropen.com/articles/10.1186/s40537-019-0197-0), and we use it the way the field uses it. What augmentation cannot do is invent coverage that was never there. Twenty photographs spread across four materials, three properties and three lighting conditions are worth more than two hundred taken on one afternoon in one bathroom. Turning the two hundred upside down does not fix them. Spread beats volume, and that is a discipline of capture, not of code. WHAT IT LEARNS FROM ### Open science, verified in the lab. Read it left to right: where the images come from, what happens to them, and the two places the finished vision runs. The lower path is the one that keeps everybody honest. Each instrument learned from the source built for it: open datasets such as OpenFungi taught the binary detector, and the dedicated genus collection from our laboratory partner taught the genus classifier. Both pass through augmentation and training. The dashed path along the bottom is the yardstick. Images whose genus a laboratory confirmed are held back from teaching so that they can mark the homework instead. PUBLIC TECHNIQUES, LAYERED WITH PROPRIETARY REFINEMENTS DEVELOPED WITH LABORATORY PARTNERS. THE SPECIFICS STAY IN THE LAB. The corpus is layered, and the two instruments learned from different parts of it. The quick mould or no mould detectors learned with the help of open datasets such as [OpenFungi](https://pmc.ncbi.nlm.nih.gov/articles/PMC12300142/), which publishes macroscopic and microscopic images of fungal genera including Aspergillus and Penicillium. The genus classifier learned from a dedicated collection of genus-labelled specimen photographs, developed in collaboration with Joseph Langford, Principal Scientist, Civil & Domestic Environmental Sciences. Field photographs contributed by trade professionals under a structured capture programme broaden both. And laboratory verified specimens alone are allowed to test what either instrument learned. That last distinction is the one that matters most, and it is the one most often skipped. A photograph can do two entirely different jobs. It can teach, which almost any reasonable image can do. Or it can mark, which means testing whether the software actually got it right. If the answer key was itself a judgement made by eye, the test measures agreement with that judgement rather than correctness, which is marking your own homework with your own answers. Anything can help it learn. Only laboratory confirmed images can tell you whether it learned the right thing. Every training image carries its context: surface, room, distance, light. That is a capture discipline rather than a clever algorithm, and it exists for one reason. It lets us audit what the model is actually learning from, and it makes the coverage grid checkable rather than a feeling. Genus down the side, surface along the top: the job is to fill it in far enough that no row can be identified by its column. Coverage also has to be local. Models trained on whatever the open web offers are overwhelmingly American, and the British and Australian building fabric is a different visual vocabulary: Artex ceilings, uPVC window reveals, paper faced plasterboard, timber skirting and coving. A classifier that has never seen a textured ceiling is guessing the moment it meets one. And then the least intuitive part of the whole exercise. A model that has never been shown a clean ceiling cannot tell you a ceiling is clean. It can only tell you which genus the ceiling most resembles, and it will do it with every appearance of confidence. So clean surfaces are collected with the same care as the mould, along with the confusables: water staining, soot and thermal ghosting, efflorescence, dust, rust, cobwebs, blown plaster. A photograph of a clean wall is not a wasted photograph. On most jobs, clean is the answer the client is hoping for, so it needs to be an answer the software can actually give. INSIDE THE CLASSIFIER ### From 1,024 clues to one clear answer. The left column is what the network noticed in your photograph. Each column after it narrows the possibilities. The right column is its answer, and one lit path shows a single decision travelling the whole way. Four columns, left to right. The first holds the 1,024 things the vision noticed in your photograph, drawn as ten nodes with an ellipsis for the rest. The next two narrow that down. The last is the answer: seven named outcomes, each with a confidence. The lit line follows one decision the whole way through, from a single clue to Penicillium. The first pass is done by a backbone, a general purpose vision network that has already learned, on millions of ordinary photographs, what an edge is, what a texture is, what a boundary between two materials looks like. We use a family of lightweight convolutional neural network architectures developed by Google, specifically the networks that [Howard and colleagues published in 2017](https://arxiv.org/abs/1704.04861), designed to run inside a phone rather than a data centre. Reusing that general sight and teaching only the last part of the network a new question is called transfer learning, and it is why a specialist classifier does not need the internet's worth of photographs to become useful. What the backbone hands over is 1,024 numbers describing your photograph. Not pixels: descriptions. From there the head of the classifier is a panel of specialists. The first layer of 256 units reads all 1,024 clues and each unit votes on a pattern it cares about. A layer of 64 listens to those votes and narrows again. The last layer holds exactly seven opinions, one per possible answer, and softmax turns them into a final show of hands: seven confidences that add up to one whole, with the tallest as the answer. Two of the head's steps exist purely to keep the panel honest. Batch normalisation keeps the numbers flowing between layers in a steady range so training does not lurch. Dropout switches off a random handful of the specialists on every training round, so the network can never come to depend on one loud voice. It has to learn an answer that survives whoever happens to be absent. A confidence is only worth having if it is honest, and the field has known since [Guo and colleagues at ICML 2017](https://arxiv.org/abs/1706.04599) that modern networks tend to state their answers more forcefully than the evidence deserves. So calibration is treated as work in its own right, not a by-product. The system is also allowed a third verdict: it can answer that it cannot tell from this photograph and ask for a better one. And when it does report mould, it can show where in the frame the evidence sits, which is the practical descendant of the work [Zeiler and Fergus](https://arxiv.org/abs/1311.2901) began on making a network's attention visible. #### The seven answers. Acremonium. Slower growing, and associated with long term moisture. Aspergillus. A large, ubiquitous genus, common on damp building materials. Chaetomium. Favours chronically wet paper faced materials such as plasterboard. Cladosporium. Among the most common moulds indoors and outdoors, and a regular on damp walls and window reveals. Penicillium. Widespread indoors, and the genus behind penicillin. Stachybotrys. The slow growing genus on saturated cellulose materials, known to the public as black mould. No mould. The seventh answer, and on most jobs the one the client is hoping for. It is only available because clean surfaces were collected as deliberately as the growth. TWO INSTRUMENTS ### One decides yes or no, the other decides what the likely risk factors are. Mould Detect runs two types of AI Vision, and the division of labour is deliberate. The first is a binary detector: one question, mould or no mould, answered fast. It is the triage instrument, built to be quick and cautious, and it runs wherever the job needs it, in the app in your hand or in the cloud. In a loft with no signal, the read can happen on the handset, and the photograph does not have to leave it. The second is the specialist this page has been describing: the genus classifier, the multi-classifier that picks up where the detector leaves off and narrows a photograph to one of six genera, or no mould, with a confidence attached. One instrument answers yes or no. The other names what you are likely dealing with. Neither is a lesser copy of the other, and the pairing is the point. Speed first, depth second, and an honest answer either way: if a photograph cannot support a verdict, the system is allowed to say so and ask for a better one. And every model we serve has to prove it behaves like the approved model before it answers anybody. GLOSSARY ### The words, plainly. Every term this page uses, in one sentence each. Nothing here needs a computing background. Artificial intelligence (AI) Software that learns its behaviour from examples rather than following rules a person wrote out by hand. Augmentation Making many training examples out of one photograph by turning, flipping, cropping and shifting it, so the model learns the mould rather than the photograph. Backbone The general purpose vision network that does the first pass of looking, before a small head is trained on one specific question. Batch normalisation A step between layers that keeps the numbers flowing through the network in a steady range, so training stays stable. Calibration The work of making a stated confidence mean what it says, so that answers given at eight in ten confidence are right about eight times in ten. Classifier The part of the system that sorts a photograph into one of a fixed set of answers. CNN (convolutional neural network) A network that looks at small patches of an image at a time and builds up from edges and textures to whole shapes. Confidence The strength the model attaches to its own answer, reported next to the answer rather than hidden behind it. Confusables Things that look like mould in a photograph and are routinely mistaken for it, such as water staining, soot, dust and salt deposits. Coverage The spread of a photograph collection across surfaces, rooms, scales and lighting conditions, which matters more than the raw number of photographs. Dropout A training habit that switches off a random part of the network each round, so no single path can come to dominate the answer. Efflorescence The white salt bloom that appears on masonry and plaster as water moves through the material and evaporates, and a classic mould lookalike. Embedding The list of numbers a network produces to describe an image, arranged so that things which look alike end up with similar numbers. False negative A miss: real mould that the system failed to report. False positive A false alarm: something reported as mould that was not mould. Genus and species The two levels of a biological name, where the genus is the family surname (Aspergillus) and the species is the first name (fumigatus). Ground truth The answer a system is measured against, and the reason it matters a great deal who or what supplied that answer. Inference Running an already trained model on a new photograph to get an answer, as opposed to training it. Microscopy Examining a sample under a microscope, which is how many species level identifications are actually settled. Mycology The study of fungi, including the moulds that grow in buildings. On-device processing Analysis that happens on the phone itself, so it is instant, works without a signal, and the photograph need not leave the handset. Shortcut learning When a model latches onto something easy that happens to predict the answer, such as the room it was photographed in, instead of the thing you meant it to learn. Softmax The final step that turns the network's raw scores into a set of confidences that add up to one whole. Substrate The material the mould is growing on: plasterboard, grout, timber, fabric, paper. Training Showing a model example after example with the right answer attached, adjusting it slightly each time so that it would have been less wrong. Transfer learning Starting from a network that has already learned to see general shapes and textures, then teaching it one new question of your own. Validation set Photographs held back from training and used only to check what was learned. HOW WE HELP Mould Detect gives you tools, information and support for confident decisions. If you suspect a mould issue, use our apps to understand and document what you are seeing, and always seek professional advice. Where possible, confirm with laboratory testing to fully understand your situation. [Join the early access](https://moulddetect.ai/#early-access) ### See what your home is growing. Early access is opening now. Point a phone at the wall, get the genus and the confidence, and keep the record. [Join the early access](https://moulddetect.ai/#early-access) --- # Mould Detect for Developers | Mould, Environmental and Weather APIs URL: https://moulddetect.ai/developers/ FOR DEVELOPERS ## Mould and weather data, one request away. The same data platform behind this site's live conditions is a documented HTTPS API at data.moulddetect.ai: mould growth index, weather history and forecasts, river flood, air quality and site layers. History, flood and air quality cover Australia; forecasts are global. Every response carries its units and its sources. THE ENDPOINTS ### One key, one envelope, every endpoint. Each row is a GET endpoint on data.moulddetect.ai. Every endpoint needs a key except /health. Mould Detect data API endpoints Endpoint What it returns /v1/mould-index A daily 0 to 6 mould growth index from 3-hourly temperature and humidity: recent ERA5-Land history joined to the GFS forecast. /v1/daily Daily aggregates for any point: up to 7 past days, today and up to 7 forecast days. /v1/now One lightweight current reading plus today's high and low, built for polling clients. /v1/history Historical daily weather from the ERA5-Land reanalysis, including soil-moisture layers. /v1/flood River discharge forecast and flood severity from GloFAS, snapped to the nearest river cell. /v1/air CAMS air quality: seven pollutants three-hourly, up to 5 days ahead. /v1/site Static site layers: elevation, slope, aspect, wetness index, soil, land cover, distance to water. /v1/hazard Fire and tropical-cyclone hazard history near a point. /v1/astro Sun and moon times for any point. /v1/specialists A directory of Australian mould and building specialists by suburb or region. /health Liveness and data freshness; check it before you request history. ENVIRONMENT.MOULDDETECT.AI/DOCS · VERSIONED UNDER /v1 A REAL REQUEST ### Two days of air quality for Sydney. Copy it, put your key in the variable, run it. The response names its model run and carries the Copernicus attribution you must show. SHELL · FROM THE PUBLISHED API REFERENCE curl -s -H "x-api-key: $MOULD_DETECT_KEY" \ "https://data.moulddetect.ai/v1/air?latitude=-33.8688&longitude=151.2093&days=2" Keys are issued per account and sent in the x-api-key header. Errors share one JSON envelope on every endpoint. Ask /health for the latest complete history date before you request /v1/history. [FULL API REFERENCE ](https://environment.moulddetect.ai/docs/)[DATA.MOULDDETECT.AI ](https://data.moulddetect.ai/health) AGENT FIRST ### The same data for your code and for your agents. Four ways in, each with its status today. - Live #### Environmental data API. The endpoints above, for apps, dashboards and back-end jobs. [API REFERENCE ](https://environment.moulddetect.ai/docs/) - Live #### Weather data. Forecasts, current conditions and reanalysis history in the same envelope, metric or imperial. [API REFERENCE ](https://environment.moulddetect.ai/docs/) - Built #### Mould detection. The AI Vision model behind the app, available with developer key access. [AI VISION ](https://moulddetect.ai/ai-vision/) - Beta #### AI agents. Mycelium, an MCP server at mcp.moulddetect.ai with read-only tools for mould risk, weather, flood, air quality and photo checks, plus llms.txt for any model. [AGENTS AND MCP ](https://moulddetect.ai/agents/) BEHIND THE API ### Where the numbers come from. Sources on the left feed the data lake; the endpoints read from it. Filled phases of the Mould Atlas run today. - Independent environmental feeds. Live ERA5-Land climate history, NOAA GFS forecasts, GloFAS river flood and CAMS air quality, collected for Australia today. - Readings from inside buildings. Prototype Sensor boards post to the same lake on a sensor-agnostic schema: a new sensor adds a key, not a new pipeline. - Scans, assessments and lab results. Collecting AI Vision scans, Mould Assessment reports and partner lab results become the labels the Atlas learns from. - Phases 0 and 1 run today. Live The data lake, and a published VTT mould-growth model per location served as /v1/mould-index. It is driven by outdoor conditions and has not yet been validated against indoor damage. - Phases 2 to 4 are the roadmap. Roadmap Learn the outdoor-to-indoor transfer by building type, train on verified outcomes, then predict risk by building type, postcode, climate zone and season. Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ### Read the reference, then build. The full reference has every parameter, response field and error. How to get an API key is covered in the docs too. [Open the API reference](https://environment.moulddetect.ai/docs/) --- # Mould Questions, Answered | Mould Detect FAQ URL: https://moulddetect.ai/faq/ FAQ ## Twenty questions people actually ask. Straight answers on mould and health, identification, and natural treatment, for homeowners, renters, landlords, cleaners and remediation specialists. HOW WE HELP Mould Detect gives you tools, information and support for confident decisions. If you suspect a mould issue, use our apps to understand and document what you are seeing, and always seek professional advice. Where possible, confirm with laboratory testing to fully understand your situation. [Join the early access](https://moulddetect.ai/#early-access) MOULD AND YOUR HEALTH ### What mould can do to you. #### Is mould dangerous to my health? Yes. Mould exposure has been linked to respiratory irritation, allergies, asthma aggravation and long-term health complications. Children, the elderly and people with weakened immune systems are most vulnerable. #### What is black mould and why is it dangerous? Stachybotrys chartarum, known as black mould, releases mycotoxins linked to chronic fatigue, severe respiratory problems and neurological issues. It thrives on water-damaged, cellulose-rich materials such as plasterboard and timber. #### What are the symptoms of mould exposure? Common symptoms include coughing, wheezing, asthma attacks, eye irritation, skin rashes, nasal congestion and throat irritation. These can mimic allergies or flu-like symptoms, which makes mould easy to overlook. #### Which types of mould are most harmful? The five most concerning types are Stachybotrys chartarum (black mould), Aspergillus, Cladosporium, Penicillium and Fusarium. Each produces different health effects, ranging from allergic reactions to serious respiratory infections. #### Can mould affect my pets? Yes. Pets can experience respiratory issues, lethargy and allergic reactions from mould exposure. Small animals and birds are particularly sensitive to airborne mould spores. #### How quickly can mould affect my health? Some people experience symptoms within hours of exposure, while others develop issues over weeks or months of chronic exposure. If symptoms improve when you leave the building, mould may be the cause. [Go deeper: the mould species guide.](https://moulddetect.ai/learn/#the-species) IDENTIFYING MOULD ### Learn to spot it early. #### How do I know if I have mould? Look for visible discolouration on walls, ceilings and corners, often appearing as black, green or white patches. A persistent musty smell is one of the earliest indicators, even when mould is not yet visible. #### Where does mould commonly grow? Mould thrives in damp, poorly ventilated areas: bathrooms, kitchens, basements, laundry rooms, around windows with condensation, behind furniture against external walls, and in roof spaces with leaks. #### What does mould look like? Mould can appear as fuzzy, slimy or powdery patches in black, green, white, orange or brown. It often starts as small spots that spread outward. Early-stage mould can look like dirt or staining. #### How can I tell the difference between mould and dirt? Apply a drop of household bleach to the spot as a test, but do not use bleach to clean mould. If the spot lightens after a few minutes, it is likely mould. Dirt will not change colour. This test damages surfaces, so use it sparingly. #### Can mould be hidden inside walls? Yes. Mould can grow behind wallpaper, inside wall cavities, under carpets and in ceiling spaces. Signs include musty odours, unexplained health symptoms, and bubbling or peeling paint. #### How fast does mould grow? Mould can begin growing within 24 to 48 hours once it has what it needs: humidity above about 60%, and dust or another organic surface to feed on. Penicillium is particularly fast-growing. This is why addressing water damage immediately is critical. [Go deeper: the safe cleaning guide.](https://moulddetect.ai/learn/#safe-cleaning) NATURAL TREATMENT AND REMEDIATION ### Deal with it, naturally. #### Can I clean mould myself? Yes, for small areas under 1 square metre. Wear a P2 respirator mask, rubber gloves and safety glasses. Use damp cloths or microfibre to trap spores, and never dry brush or vacuum dry mould. #### What natural solutions kill mould? White vinegar, at 3 parts vinegar to 1 part water, is effective against most mould species. Tea tree oil, at 1 teaspoon per cup of water, is also effective and helps prevent regrowth. Both are non-toxic alternatives to chemical cleaners. #### How do I prevent mould from coming back? Address the moisture source first: fix leaks, improve ventilation and keep indoor humidity below 60%. Open windows daily, use exhaust fans, keep furniture away from walls, and wipe condensation from surfaces regularly. #### Should I throw away mouldy items? Porous items such as cushions, mattresses, carpets and soft furnishings that are heavily contaminated usually cannot be cleaned effectively and should be disposed of. Hard, non-porous surfaces can be cleaned. #### Is vinegar better than bleach for mould? For most household mould, yes. Vinegar penetrates porous surfaces and kills mould at the root. Bleach only cosmetically removes surface mould on non-porous materials, and because it contains water it can cause mould to return and can damage surfaces. Vinegar and natural tree oil based solutions can be better alternatives. #### When should I call a professional? Call a professional if mould covers more than 1 square metre, if it returns quickly after cleaning (which indicates structural dampness), if it is in heating or air-conditioning systems, or if anyone in the household has serious health symptoms. #### How do I improve ventilation to prevent mould? Open windows for at least 15 minutes daily, use exhaust fans in bathrooms and kitchens during and after use, ensure dryer vents exhaust outside, and consider a dehumidifier for persistently damp rooms. #### Can indoor plants cause mould? Overwatered indoor plants can add humidity and grow mould in the soil and the area around them. If you already have a mould problem, reduce the number of indoor plants and avoid overwatering. [Go deeper: the natural solutions guide.](https://moulddetect.ai/learn/#natural-solutions) ### Still unsure what you are looking at? Scan it, keep the record, and get a first opinion in seconds. Mould Detect is opening to early access now. [Join the early access](https://moulddetect.ai/#early-access) --- # For Specialists | Mould Detect URL: https://moulddetect.ai/for-specialists/ FOR SPECIALISTS ## A sixty-second scan. A report that holds. Built for inspectors, remediators, builders, plumbers, cleaners and environmental health professionals. Scan on site, generate a report in minutes, and route confirmed cases straight to remediation or to a laboratory. THE FILM · 2:42 ### One photo becomes a report under your own name. Miss Mould, our AI mycologist, walks one site visit from the first photo to the finished report. Captions are on by default. Your browser cannot play this video. [Download the MP4](https://moulddetect.ai/video/mould-detect-for-professionals/mould-detect-for-professionals.mp4). 2:42 · ENGLISH CAPTIONS · SCAN RESULTS SHOWN ARE SAMPLES NEW REVENUE ### Turn every job into a mould assessment line on the invoice. Three ways the same scan earns for a specialist, residential or commercial. - #### Indicate the risk, with confidence. Give residential and commercial customers a mould risk indication with a confidence score and the local conditions behind it, on the first visit and before you quote. - #### Remediate it yourself, or with our help. Treat the job yourself with documented before-and-after scans, or route it through Mould Detect to laboratory testing with approved partners and to remediation specialists. - #### Point customers to what fixes it. Direct customers to the products and services that remediate and prevent mould, from ventilation to treatment, so the job does not end at the diagnosis. ENVIRONMENTAL DATA IN THE APP ### Know the conditions that are driving the damp. The cards give the conditions at a property now; the chart shows how long humidity has sat above the 70% damp line, which says more than any single reading. Every reading shown here is a sample. The app's environment view for specialists: weather, humidity and air quality Why it helps your business - Explain the cause. Show a client how long humidity sat above the damp line, so they see why the mould grew. - Plan the work. The Ventilation Planner shows when opening the windows helps and when to keep them shut. - Keep it together. The scan result and local conditions sit side by side in one app. - Weather Light rain - Temperature 18 °C - Humidity 82% · Damp - Air quality PM2.5 · Good Sample readings Air quality: Generated using Copernicus Atmosphere Monitoring Service information 2026 Humidity history Sample humidity history: 12 daily readings from 62% to 82%, at or above the 70% damp line on 8 of them. Mould Detect gives you more to decide with than the photo: local weather, temperature, humidity history and air quality for the property, from the platform behind our public API. Use it to judge whether the growth will come back after treatment, and to explain the cause to your client. Coverage is Australia. THE WORKFLOW DETECT · DOCUMENT · DESTROY ### One process, from the first walk-through to the final invoice. - DETECT #### Scan on site. Read the risk areas with the phone already in your pocket. AI Vision gives you a genus-level read and a confidence score before the walk-through is finished, on-site and without waiting on a lab. - DOCUMENT #### A white-label report. Every scan becomes part of a report under your own name: timestamped, location-tagged and ready to hand to a client, an insurer or a strata manager. - DESTROY #### Route to remediation, or to a lab. Confirmed cases route straight to remediation. Where the read calls for it, they route on to laboratory testing instead. Post-treatment scans supply the before-and-after evidence the job needs. DEFENSIBILITY & REPORTING ### Reports that hold up when someone asks questions. White-label PDFs. Reports go out under your own name, not ours. Available from Business and up. Pre and post-treatment evidence. Document the wall before you start and again after you finish, so the before-and-after speaks for itself. Fewer callbacks. A documented on-site assessment before quoting catches what a walk-through alone can miss. Higher invoice values. A defensible report is evidence of the work done, which is easier to justify and easier to bill. Documentation standards are rising on both sides of the Tasman and beyond: Awaab's Law in the UK and the NCC 2025 condensation management provisions in Australia both point the same way, toward assessments that are timestamped, evidenced and easy to produce on request. DATA & ANALYSIS ### The live conditions platform, in your workflow. Walk in already knowing the conditions. The environmental platform behind the app's live conditions feeds every inspection with the local context that explains what you find on the wall. Current weather and humidity. Local relative-humidity trends and temperature at the property give a damp reading its context: whether today is unusual, or the building is always working against its climate. Regional flood risk. Flood severity history for the region flags properties where past water events make hidden moisture the likely story behind a scan. Air quality and ventilation. Air-quality feeds drive ventilation guidance: when opening windows helps a drying plan, and when it works against it. Moisture and condensation outlooks. Condensation-risk outlooks show the days ahead when a treated room is most likely to relapse, so your recommendations land with dates attached. Professional accounts see this context in the app for every property on their books. Enterprise adds API access and insurer-ready exports, so the same insight sits inside your own systems rather than a separate app. Decision support only. Mould Detect is a decision-support tool, not a diagnostic instrument or a substitute for laboratory confirmation. It is built to sit alongside your professional judgement, and to speed up what you already do, not to replace it. OCCUPATIONAL HYGIENISTS & LABORATORIES ### Ecosystem partners, not competitors. A chain of evidence, not a photo roll. Every location-tagged scan carries a timestamp and a confidence score, building a record a laboratory or hygienist can pick up rather than start over. A prompt to confirm, not a diagnosis. A likely-positive read routes straight to laboratory testing. Mould Detect never claims to replace what only a lab can confirm. Enterprise API access. Labs and OH practices pull scan data directly into their own systems, at insurer-ready quality. SPECIALIST TIERS ### Three tiers, one ladder. - Business $49 /month Small and medium firms running an on-site assessment before quoting. - Professional $199 /month Remediation firms, franchises and inspection practices running a team across many properties. - Enterprise $399+ /month Labs, insurers and mycological partners who need API access and insurer-ready data. [See the full pricing matrix, feature by feature.](https://moulddetect.ai/pricing/) Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ### Put it to work on your next job. Become a Mould Detect partner. Contact us for a demo and next steps, and start scanning, documenting and routing with confidence. [Contact us for a demo](https://moulddetect.aimailto:hello@moulddetect.ai?subject=Mould%20Detect%20partner%20demo) --- # Mould Detect for Homeowners and Renters | Fast Answers, Safe Next Steps URL: https://moulddetect.ai/homes/ FOR HOMEOWNERS AND RENTERS ## Know what is growing on your wall. Point your phone at a suspect patch and Mould Detect gives you a plain-language read on the mould risk in under 60 seconds, with what to do next. That is far quicker than waiting on a laboratory. It does not replace one: when you need certainty, the app points you to lab testing with approved partners. SCAN FIRST, LAB WHEN IT MATTERS ### A first answer in a minute. A lab answer when you need one. Read across each row: how long it takes, what it costs, and what you come away with. A Mould Detect scan compared with laboratory testing Time to a result Cost What you get Mould Detect scan Under 60 seconds Free for 10 scans a month; paid plans from $19.49/mo A likelihood indication, a confidence score and a recommended next step Laboratory test 7 to 14 days $300 to $1,800+ The species, confirmed by a laboratory Use the scan to decide whether a lab test is worth it, to keep a dated record, and to act sooner. A scan is decision support only. It is not a lab test and not medical advice. MOULD DETECT PITCH DECK · 14 OCT 2026 · PRICING AT MOULDDETECT.AI/PRICING ONE PHOTO ### One clear photo, one clear verdict, and why. The clip shows a real scan in the app. The list beside it is what you get back, in the order you will read it. The app analysing a photo of a suspect patch - How likely it is mould. High, medium, low or unlikely: a visual signal, never a lab diagnosis. - How sure the AI is. A confidence score, and a heat map of where it looked, so you know when to re-scan. - What to do next. Monitor it, clean it safely, book a professional, or order a lab test. FAMILY SAFETY ### Peace of mind for the people who live there. Four things the app does for a household, and where to start each one. - #### Know which days to open the windows. Live humidity and air quality for your city, with a planner that says which hours are worth ventilating. Occupants of damp or mouldy buildings face up to a 75% greater risk of respiratory symptoms and asthma (WHO, 2009), and drying the air is the first defence. [LIVE CONDITIONS ](https://moulddetect.ai/#live-conditions) - #### Evidence for a landlord conversation. Every scan is dated and kept, so a renter can show what was on the wall and when, rather than describe it from memory. - #### Plain answers, with sources. Ask Miss Mould, the AI mycologist in the app, and she answers in plain language and shows where the answer came from. [READ THE FAQ ](https://moulddetect.ai/faq/) - #### Clean it safely, or know when not to. Mould Edu explains what you can clean yourself, how to do it without spreading spores, and when a professional should do it instead. [MOULD EDU ](https://moulddetect.ai/learn/) WHEN YOU NEED CERTAINTY ### A likely-positive scan points you to a laboratory. The clip shows the app's next step after a likely-positive result: lab testing with an approved partner. The app recommending lab testing with an approved partner Only a laboratory can confirm which mould is present. Mould Detect helps you decide when that test is worth paying for, and vetted remediation specialists near you are coming to the app. Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ### Start with a free scan. Join the early access and scan your first suspect patch the day the app opens to you. Free stays free. [Get early access](https://moulddetect.ai/#early-access)[Compare plans](https://moulddetect.ai/pricing/) --- # Mould Detect | AI Mould Detection for Australian Homes URL: https://moulddetect.ai/ AI-Powered Mould Intelligence · Australia ## See what your home is growing. Point your camera at a suspect wall. In sixty seconds, Mould Detect reads the risk, explains it in plain language, and shows you what to do next. [Get early access](https://moulddetect.ai/#early-access)[How it works](https://moulddetect.ai/#how-it-works) SYDNEY · 11:09 · CONDITIONS: READING - [ ](https://www.linkedin.com/posts/theventures-australianstartups-startupawards-share-7518987186017333248-pfw4/) - [ ](https://www.linkedin.com/posts/theventures-australianstartups-startupawards-share-7518987186017333248-pfw4/) 1 in 2 Australian homes report mould or dampness $28B Annual asthma-related health costs in Australia 60 sec From photo to plain-language risk read Launching in Australia to prove the approach · the UK next, then Canada and the US Find your door ### Go straight to the page written for you. Eight doors in three groups. Pick the one that describes you; every page links back to the others. Meet Mould Detect, with Miss Mould · 2:02 · English captions People and homes - [ Homes & renters A quick read on a suspect patch, and what to do next. ](https://moulddetect.ai/homes/) - [ Specialists & pros Assess mould risk on site and turn it into new work. ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers Independent environmental data and evidence, before the claim. ](https://moulddetect.ai/insurers/) - [ Investors The market, the data moat and the team. ](https://moulddetect.ai/investors/) - [ Partners Trades, inspectors, hygienists and labs who build the dataset. ](https://moulddetect.ai/partners/) Builders of technology - [ Developers Mould, environmental and weather data over a documented API. ](https://moulddetect.ai/developers/) - [ AI agents An MCP server and llms.txt, built agent first. ](https://moulddetect.ai/agents/) - [ IoT & devices Sensor boards that post into the same data lake. ](https://moulddetect.ai/iot/) The process ### AI enabled insights in your pocket. DETECT · DOCUMENT · DESTROY - #### Scan Open the app and frame the area. Our vision model reads texture, spread and staining that a checklist can't. - #### Understand A calm, plain-language read of what's growing, how serious it is, and what's feeding it. Ask the mycology AI anything. - #### Resolve Reports formatted for landlords and insurers. Vetted remediators and natural, non-toxic treatment guidance when you need them. From photo to action #### Your phone. Sixty seconds. A next step you can act on. One photo becomes four readings: likelihood, risk rating, confidence score and a recommended next step. Every reading opens a path you can take immediately: learn, ask the AI, order a lab test, or book a certified remediator. Inside the app #### An AI that shows its working. Two things you will not find elsewhere: computer vision that shows its evidence, and an expert you can question until the answer makes sense. See where the AI looked. Every scan comes with an attention map showing exactly which parts of your photo drove the reading. No black box: you can check the evidence against your own eyes. SALIENCY HEATMAP · FROM A REAL SCAN Ask until you understand. Miss Mould, our AI mycologist, answers in plain language from a reviewed knowledge base and shows her sources. In the web app, share your location and she can also explain today's weather and this week's mould risk for your area, naming where the numbers come from. It is decision support from outdoor conditions, not an inspection: she tells you when a professional should take over. MISS MOULD AI · IN THE APP Know which genus you are dealing with. Thousands of mould species exist, but only a fraction are allergenic, pathogenic or toxigenic. Our experimental Mould Genus feature suggests which genera your photo is most consistent with, so you can judge urgency sooner; laboratory testing then confirms it. Developed with the kind support of Joseph Langford, Principal Scientist, Civil and Domestic Environmental Sciences. MOULD GENUS PANEL · EXPERIMENTAL Built · Shipped · Live #### The ecosystem behind every sixty-second read. Five AI models, a live environmental data platform and three shipping applications feed one AI Vision engine, hosted in AWS and built with Cloud Assembly through the AWS Partner Program Startup Accelerator. Miss Mould AI reads local weather and mould risk from the same platform, through our Mycelium MCP server, when you share your location. AI Vision training data sources - 1,249 high-quality macroscopic and microscopic images from the OpenFungi dataset - 50,000+ training images - 6 mould genera, lab verified Who it's for ### Built for everyone who lives with walls. - FAMILY HEALTH · TENANCY EVIDENCE #### Homeowners & Renters Peace of mind in your pocket. Scan the risk areas of your home, know what your household is exposed to, and document evidence to improve discussions with landlords. [JOIN THE EARLY ACCESS](https://moulddetect.ai/#early-access) - MULTI-PROPERTY MANAGEMENT #### Specialists Rapid triage, reporting and workflow for inspectors, remediators, builders, cleaners, plumbers and environmental health experts. [BUILT FOR YOUR WORKFLOW](https://moulddetect.ai/for-specialists/) - SITE ASSESSMENT · CHAIN OF EVIDENCE #### Occupational Hygienists & Laboratories Consistent, location-tagged capture in the field. Structured records that support sampling plans, laboratory analysis and defensible reporting. [ECOSYSTEM PARTNERS](https://moulddetect.ai/for-specialists/#oh-and-labs) [COMPARE ALL TIERS · PRICING](https://moulddetect.ai/pricing/) Why Mould Detect ### Cutting-edge AI vision, trained on tens of thousands of images. Macro footage zooming into a mould colony until individual spore structures are visible. Early risk assessment. A lab test can take days or weeks; understand the risks sooner. Mould Detect is not a replacement for lab testing, and we will always recommend following up with a lab test. Trained on real world data. Computer vision built in collaboration with mycological labs and mould remediation professionals. Documentation that stands up. Reports formatted for cleaners, building inspectors, mould remediation specialists, property managers, landlords and insurers. Natural and kind to the environment. Guidance favours proactive and preventative natural, non-toxic solutions and treatment first. AI VISION ### A neural net trained to detect details the human eye often does not notice. The diagram illustrates the activation sequence of neurons across layers of the Convolutional Neural Network (CNN). Mould Detect runs two kinds of AI Vision: a fast detector that answers mould or no mould, and a genus classifier that names which of six genera it is looking at. They learned from open datasets such as OpenFungi and a dedicated genus collection built with our lab partner, and only laboratory-confirmed images are allowed to test them. The answer arrives on the spot: which genus, and how confident. [How our AI sees mould](https://moulddetect.ai/ai-vision/)[Join the early access](https://moulddetect.ai/#early-access) AI AND AGENT FIRST ### Our data is built for AI agents as well as people. Agent First is how Mould Detect now builds its APIs and data infrastructure. Each data service we publish, from the mould growth index to weather history, flood forecasts, air quality and the photo check, is also a tool an AI agent can find, call and cite. Mycelium Network is that layer: one MCP server for Claude, ChatGPT, Cursor and your own agents, a terminal agent called Mycelium CLI, and docs written for machines as well as developers. Every answer carries its sources and licence notices. [Mould Detect for agents](https://moulddetect.ai/agents/)[MYCELIUM.MOULDDETECT.AI →](https://mycelium.moulddetect.ai/) Mycelium CLI The data platform ### Growing conditions, right now. Mould does not wait for an inspection. This is what our models read for your city: the seven-day risk outlook, the humidity trend behind it, and the hours worth opening a window, drawn from the Mould Detect data platform. LIVE CONDITIONS UNAVAILABLE RIGHT NOW · THE FEED RETURNS AUTOMATICALLY SEVEN DAY OUTLOOK #### Know which days need your attention. Each bar counts the hours the outside air sits above 80 percent humidity. Indoors, mould needs about 60 percent humidity, dust to feed on and 48 hours, and long damp stretches outside are when cold walls and cupboards cross that line. Plan drying, airing and checks around the labelled days. VIEW AS TABLE HUMIDITY · 3-HOURLY #### Watch the number mould actually responds to. VIEW AS TABLE of humidity readings GROWTH CONDITIONS FIELD #### See how close your city is to growth conditions. VIEW AS TABLE of growth regions VENTILATION PLANNER #### Open windows when it helps, not when it hurts. VIEW AS TABLE of ventilation slots ### Know before it grows. Early access is opening city by city across Australia. Leave your email and be first through the door. --- # Mould Detect for Insurers | Mould Risk Data, Evidence and Prediction URL: https://moulddetect.ai/insurers/ FOR INSURERS ## See mould before it becomes a claim. Insurers met mould once, in Texas, and answered by excluding it. Mould Detect collects the evidence that makes the risk visible early: independent environmental data, readings from inside buildings, photo scans and specialist assessments, in one data lake. We are launching in Australia to prove the approach and the technology, then plan to expand to the UK, Canada and the US. Insurer partnerships are an opportunity we are building toward; there is no insurer pilot or contract today. MOULD IS THE NEW ASBESTOS ### In Texas, mould claims outran the cover in three years. Read left to right: how fast Texas mould claims grew, the verdict that made the headlines, and how the industry responded. The comparison with asbestos is about claims exposure. It makes no health comparison. $420M → $2B+ Texas mould-related insurance claims, 2000 to 2002 BUSINESS PLAN v3.1 · 14 OCT 2026 $32M jury verdict in a single case, Ballard v. Farmers BUSINESS PLAN v3.1 · 14 OCT 2026 May 2002 ISO's Fungi or Bacteria Exclusion: cover withdrawn, risk unchanged BUSINESS PLAN v3.1 · 14 OCT 2026 THE RISK DID NOT LEAVE WITH THE COVER · IT CAN STILL BECOME A LONG-TAIL CLAIMS PROBLEM THE HEALTH AND COST PICTURE ### Damp buildings cost health, and health costs money. Four facts that size the problem, each with its source. - #### Up to 75% greater risk of respiratory symptoms and asthma. The WHO's 2009 dampness and mould guidelines, written by 36 experts reviewing more than 100 studies, put the risk for occupants of damp or mouldy buildings at up to 75% higher, and found that preventing or fixing dampness reduces the harm. - #### 1 in 2 Australian homes report mould or damp. Asthma Australia's figure, and the reason a risk this common has to be measured rather than guessed. - #### $28B a year in asthma-related health costs. Asthma Australia's estimate for Australia, before any building or contents loss is counted. - #### Sick building syndrome. The NHS describes it as symptoms you get while you are in a particular building, from headaches and a blocked nose to coughing, wheezing and tiredness, and lists poor ventilation and poorly maintained air conditioning among the possible causes. Ventilation and humidity are what our live data reads, building by building. [NHS · SICK BUILDING SYNDROME ](https://www.nhs.uk/conditions/sick-building-syndrome/) AWAAB'S LAW · ENGLAND ### Fixed deadlines for damp and mould, measured in days. Each row is a clock that starts when a social landlord becomes aware of a hazard. Every one of them needs a timestamped record to prove it was met. Awaab's Law timeframes for social landlords in England Duty Timeframe Emergency hazards Investigate and carry out safety work within 24 hours. Significant hazards: investigate Within 10 working days of becoming aware of the hazard. Written summary to the tenant Within 3 working days of the investigation concluding. Significant hazards: safety work Complete within 5 working days of the investigation concluding. Supplementary preventative work Begin within 5 working days; if that is not possible, physically start within 12 weeks, and finish within a reasonable time. Safety work cannot be done in time Provide suitable alternative accommodation at the landlord's expense. Awaab's Law is named after Awaab Ishak, who died in 2020 aged two from a severe respiratory condition due to prolonged exposure to mould in his home, after his parents had complained repeatedly to their social landlord for three years. The requirements came into force on 27 October 2025. GOV.UK · AWAAB'S LAW: GUIDANCE FOR SOCIAL LANDLORDS, TIMEFRAMES FOR REPAIRS WHAT AWAAB'S LAW REQUIRES ### A legal duty in every social tenancy, proved by the landlord's own records. Four things an underwriter should know about the law, each linked to the official source. - #### Who it binds. Registered providers of social housing in England: local authorities and housing associations. The Social Housing (Regulation) Act 2023 writes compliance into every social tenancy as an implied term, and the Hazards in Social Housing (Prescribed Requirements) (England) Regulations 2025 set the detail. [GOV.UK GUIDANCE ](https://www.gov.uk/government/publications/awaabs-law-guidance-for-social-landlords/awaabs-law-guidance-for-social-landlords-timeframes-for-repairs-in-the-social-rented-sector) - #### Damp and mould first, more hazards to follow. From 27 October 2025: all emergency hazards, and damp and mould that present a significant risk of harm. In 2026: excess cold and heat, falls, structural collapse and explosions, fire and electrical hazards, and hygiene and food safety. In 2027: every remaining hazard in the Housing Health and Safety Rating System except overcrowding. [GOV.UK · PHASING ](https://www.gov.uk/government/publications/awaabs-law-guidance-for-social-landlords/awaabs-law-guidance-for-social-landlords-timeframes-for-repairs-in-the-social-rented-sector) - #### Records are the defence. A landlord has a defence if it can show it took all reasonable steps to comply. The guidance asks for clear records of engagement, investigations and communications. Tenants can take a landlord to court for breach, and the Housing Ombudsman can find maladministration when time limits are missed. [GOV.UK · RECORDS AND REDRESS ](https://www.gov.uk/government/publications/awaabs-law-guidance-for-social-landlords/awaabs-law-guidance-for-social-landlords-timeframes-for-repairs-in-the-social-rented-sector) - Planned #### Private rentals are next. The government's roadmap for the Renters' Rights Act 2025 says it will extend Awaab's Law to the private rented sector, with legally enforceable timeframes for private landlords. No date is set: the timing is subject to consultation. [GOV.UK · RENTERS' RIGHTS ROADMAP ](https://www.gov.uk/government/publications/renters-rights-act-2025-implementation-roadmap/implementing-the-renters-rights-act-2025-our-roadmap-for-reforming-the-private-rented-sector) WHAT IT COULD MEAN FOR INSURERS ### When the law starts a clock, dated evidence becomes part of the risk. The first two cards are our reading of the law's effect on cover; the last two are the regulation we can cite beyond England, and what it signals. None of this is legal advice. - #### Missed deadlines could become liability claims. A tenant who can sue for breach, and an ombudsman who can find maladministration, turn a slow repair into a potential claim against the landlord. Insurers of social landlords could start to ask how the clock is being kept. - #### Earlier action could mean smaller losses. Fixed timeframes push damp to be found and fixed sooner, while it is a repair and before it is a remediation. A dated scan, an assessment and the conditions behind them are the record a landlord, and its insurer, could rely on. - In force #### Scotland followed within a year, for private landlords too. The Investigation and Commencement of Repair (Scotland) Regulations 2026 came into force on 6 October 2026 for social and private landlords: investigate damp or mould within 10 working days, report within 3, and start repairs within 5. Social landlords must finish within 20 working days. [GOV.SCOT GUIDANCE ](https://www.gov.scot/publications/awaabs-law-guidance-landlords-scotland/pages/2/) - #### A direction of travel for other markets. Two UK nations putting damp and mould on a legal clock signals where housing regulation could head elsewhere. Australia's National Construction Code 2025 condensation management provisions push the same way. Mould Detect plans to follow this path from Australia to the UK, then Canada and the US; we make no forecast of when, or whether, those markets legislate. ONE COMPANY, ONE PLATFORM ### Prevent, protect, detect and correct, in one loop. Follow the ring clockwise from the top. Each step hands its result to the next, and every outcome returns to the data lake at the centre. The chips beside each step say how much of it runs today. - Prevent. Live Live humidity, air quality and a ventilation planner say when to open a window; partner ventilation and treatment products handle the rest. - Protect. Prototype Sensor boards and alerts watch a building between visits; the Mould Atlas will forecast where risk is rising. - Detect. Beta AI Vision reads one photo of a suspect surface and returns a verdict with a confidence score in under 60 seconds. - Correct. Partners Likely-positive scans route to laboratory testing with approved partners and to remediation specialists, who correct the conditions behind the damp and the growth itself; the outcome returns to the data lake. THE MOULD ATLAS ### A risk engine fed by data an insurer cannot collect alone. Sources on the left feed the data lake; the lake feeds the Mould Atlas, Mould Detect's mould risk prediction engine. Filled phases run today; outlined phases are the roadmap. - Independent environmental feeds. Live ERA5-Land climate history, NOAA GFS forecasts, GloFAS river flood and CAMS air quality, collected for Australia today. - Readings from inside buildings. Prototype Sensor boards post to the same lake on a sensor-agnostic schema: a new sensor adds a key, not a new pipeline. - Scans, assessments and lab results. Collecting AI Vision scans, Mould Assessment reports and partner lab results become the labels the Atlas learns from. - Phases 0 and 1 run today. Live The data lake, and a published VTT mould-growth model per location served as /v1/mould-index. It is driven by outdoor conditions and has not yet been validated against indoor damage. - Phases 2 to 4 are the roadmap. Roadmap Learn the outdoor-to-indoor transfer by building type, train on verified outcomes, then predict risk by building type, postcode, climate zone and season. Over time the Atlas is designed to output mould risk by building type and location, seasonal forecasts, and the trigger thresholds that automated responses depend on, with a university research partner helping to improve the models. PARAMETRIC COVER · AN OPPORTUNITY ### Independent data that could trigger cover, or prevent the claim. Read top to bottom: the independent data sources, the humidity line rising until it crosses a threshold agreed in advance, then the responses that fire automatically. - Oracles. Live Independent data no party to the contract controls: weather and flood feeds, in-building sensors, photo scans, specialist assessments and lab results. - A threshold agreed in advance. Illustrative For example relative humidity above X% for N days, a GloFAS flood exceedance, or a condensation risk from the dew-point spread. - A. Environmental control. Phase 4 Ventilation, a dehumidifier or the diffuser switched on when a trigger fires. - B. Alerts. Prototype IoT alerting to landlords, facilities and maintenance teams, early. - C. Remediation booking. Partners A lab test or an approved remediator, booked through the supplier list. - D. Data and API feed. Live The same readings delivered to partners, including insurers, into their own systems. The technical pattern only. No insurer pilot or contract exists today. Index-based (parametric) products pay or act when an independently measured index crosses an agreed threshold, so they rely on objective, verifiable data (Swiss Re Corporate Solutions). Mould Detect already collects that kind of data for Australia. An insurer could use the same readings to price cover, to trigger a payment, or to fire a prevention response before damage becomes a claim. Insurer data partnerships are later-round optionality, and no insurance revenue is in our plan. PROTOTYPE HARDWARE ### Hardware that reads the room between visits. Left, the board running today. Right, the diffuser in design. Each caption says which is which. Prototype The ESP8266 test board posts readings into the data lake today. Temperature, humidity and dew point sensors are the next build phase. Planned 2027 Concept design of the sensing diffuser, from the Ventures Australia 2026 deck: ultrasonic diffusion plus a sensor board for our environmental data. In R&D, planned as a 2027 prototype. Concept design by Rozario Chivers, Vladimir Lekic and Lucas Chivers. The schema, the sensor phases and the roadmap are on the [IoT and embedded devices page](https://moulddetect.ai/iot/). THE MOAT ### What a competitor would have to rebuild: years of data and partners. Inputs at the edges feed the lake in the centre. The wheel turns as each stage feeds the next: more scans, more labelled data, better models, more users. - Weather and climate. Live Copernicus ERA5-Land history and NOAA GFS forecasts, collected for Australia today. - Flood and air quality. Live GloFAS river-flood forecasts and CAMS air quality, in the same lake. - Building-remediation images. Partners Photographs from a remediation network with 1,500+ verified remediation projects behind it. - Lab results from partners. Collecting Lab results collated through approved remediators, cleaners and lab partners; collation is under way and becomes the highest-grade ground truth. - Specialist assessments. Beta Room-by-room reports from the Mould Assessment app, saved as structured data. AI has made software cheap to write, so an app on its own is easy to copy. The economic model that brings trades, inspectors, labs and remediators onto one platform is harder to copy, and so are the partnerships behind it and the data they accumulate: weather and environmental history, building-remediation images, and lab results from partners as collation completes. AI Vision, the model that reads a photo of a suspect surface, has been reviewed by two patent lawyers and confirmed as worthy of a patent application. WHERE WE LAUNCH ### Proven in Australia first, then the markets where the law is moving. In order: the markets we plan to enter. The UK comes next because Awaab's Law already sets the clock. - Now #### Australia: prove the approach and the technology. Australia is our testing ground. One in two homes have damp or mould, our founders' trade network is here, and our environmental data covers Australia. - Planned #### The United Kingdom, next. The UK puts a legal clock on mould: Awaab's Law, in force since 27 October 2025. Fast triage and a dated record help landlords act in time. Entry is gated on revenue. - Planned #### Canada and the US, at Series A. Canada and the US follow, after the UK. Where claims and property deals rely on accredited lab results, as in the US, a partner lab supplies the result and Mould Detect supplies the triage and referral. Sources - [GOV.UK · Awaab's Law: guidance for social landlords, timeframes for repairs in the social rented sector](https://www.gov.uk/government/publications/awaabs-law-guidance-for-social-landlords/awaabs-law-guidance-for-social-landlords-timeframes-for-repairs-in-the-social-rented-sector) - [GOV.UK · Implementing the Renters' Rights Act 2025: our roadmap for reforming the private rented sector (13 November 2025)](https://www.gov.uk/government/publications/renters-rights-act-2025-implementation-roadmap/implementing-the-renters-rights-act-2025-our-roadmap-for-reforming-the-private-rented-sector) - [Scottish Government · Awaab's Law: guidance for landlords in Scotland (September 2026)](https://www.gov.scot/publications/awaabs-law-guidance-landlords-scotland/pages/2/) - [NHS · Sick building syndrome](https://www.nhs.uk/conditions/sick-building-syndrome/) - [WHO · Guidelines for indoor air quality: dampness and mould (2009)](https://www.who.int/publications/i/item/9789289041683) - [Swiss Re Corporate Solutions · What is parametric insurance?](https://corporatesolutions.swissre.com/insights/knowledge/what_is_parametric_insurance.html) - Asthma Australia · prevalence of mould and damp in Australian homes; annual asthma-related costs · as cited in the Mould Detect pitch deck, 14 Oct 2026 - Mould Detect business plan v3.1 · 14 October 2026 · Texas claims, Ballard v. Farmers, ISO exclusion; platform, Mould Atlas and IoT statuses - Mould Detect · Ventures Australia 2026 deck, slides 20 and 39 · diffuser concept design; market sequence: Australia, then the UK, then Canada and the US Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ### Talk to us about data partnerships. If you are exploring mould prevention, evidence or parametric products, we will walk you through the data lake, the API and the Atlas roadmap. [Email the team](https://moulddetect.aimailto:hello@moulddetect.ai?subject=Insurer%20enquiry)[See the data APIs](https://moulddetect.ai/developers/) HELLO@MOULDDETECT.AI · SYDNEY, AUSTRALIA --- # Investors | Mould Detect URL: https://moulddetect.ai/investors/ INVESTORS ## The scan is the product. The data it leaves is the moat. Mould is the new asbestos: a building and health risk that half of Australian homes already carry, still assessed by $300+ lab tests that most people never order. Mould Detect replaces the first step with a 60-second scan, and turns every scan into lab-verified data no one else holds. We are launching in Australia to prove the approach and the technology, with the UK, Canada and the US to follow. The seed round is open, and the deck is available on request. THE MARKET ### The category is large, measurable and already regulated. These are the four numbers that size it. Read left to right: the prevalence, the cost it creates, the market forming around it, and the professional base Mould Detect sells to. 50% of Australian homes report mould or damp ASTHMA AUSTRALIA $28B annual asthma-related health costs in Australia ASTHMA AUSTRALIA $41.6B indoor air quality halo market, up from $29.2B GRAND VIEW RESEARCH 404,300 construction trades workers in Australia, the serviceable base JSA / ABS · FEB 2026 MOULD REMEDIATION CORE MARKET $1.23B (2023) TO $1.52B (2030) · GRAND VIEW RESEARCH · RESEARCH AND MARKETS TAM · SAM · SOM ### Sized from the workforce already working in damp rooms. Read top to bottom, from the whole market to the share a bottom-up model says is obtainable by year 3. The SOM is a forecast from a pre-revenue company, and is labelled as one. Mould Detect market sizing Measure Figure Basis TAM · indoor air quality $29.2B to $41.6B by 2030 Global IAQ solutions market, 6.1% CAGR (Research and Markets). A halo market that buys adjacent products. TAM · mould remediation (core) $1.23B (2023) to $1.52B by 2030 Mold remediation service market, 3.0% CAGR (Grand View Research). SAM · Australia, year 1 focus 404,300 trades workers Construction trades workers, ANZSCO 33 (JSA/ABS, Feb 2026), within about 797,400 building and property trades. SOM · year 3 (forecast) $15.2M to $31.3M ARR 10,000 to 18,000 paying users in a bottom-up model; all figures are forecasts. MOULD DETECT CONCISE PITCH DECK · 14 OCT 2026 WHY INSURERS WILL CARE ### Mould is the new asbestos, measured as claims exposure. Four points from the insurer page, in one line each. Insurer partnerships are later-round optionality; none is in the plan's revenue. - #### A precedent insurers remember. Texas mould claims ran from $420M in 2000 to over $2B in 2002, and the industry answered by excluding fungi in May 2002. [THE PRECEDENT ](https://moulddetect.ai/insurers/#precedent) - #### Deadlines measured in days. Awaab's Law gives social landlords in England 10 working days to investigate significant damp and mould hazards, and 24 hours for emergencies. [AWAAB'S LAW ](https://moulddetect.ai/insurers/#awaabs-law) - #### Buildings that make people unwell. The NHS lists poor ventilation among the possible causes of sick building syndrome; ventilation and humidity are what the platform reads. [HEALTH AND COST ](https://moulddetect.ai/insurers/#health-and-cost) - #### Data that could trigger cover. Independent environmental data, sensors and evidence are what parametric products need. The pattern is drawn; no insurer contract exists. [PARAMETRIC ](https://moulddetect.ai/insurers/#parametric) ONE COMPANY, ONE PLATFORM ### Prevent, protect, detect and correct, in one loop. Follow the ring clockwise from the top; every outcome returns to the data lake at the centre. The chips say how much of each step runs today. - Prevent. Live Live humidity, air quality and a ventilation planner say when to open a window; partner ventilation and treatment products handle the rest. - Protect. Prototype Sensor boards and alerts watch a building between visits; the Mould Atlas will forecast where risk is rising. - Detect. Beta AI Vision reads one photo of a suspect surface and returns a verdict with a confidence score in under 60 seconds. - Correct. Partners Likely-positive scans route to laboratory testing with approved partners and to remediation specialists, who correct the conditions behind the damp and the growth itself; the outcome returns to the data lake. THE MOAT ### Every scan makes the next one better. Detection is the product. The moat is what detection leaves behind. Every scan Mould Detect runs is a labelled observation of a real building, and the remediation network behind the company tells us what was actually there. A lab-verified dataset, not a photo archive. The training corpus is drawn from a remediation network with 1,500+ verified remediation projects behind it. Each scan can be checked against what remediation found, which is the difference between labelled data and collected data. Diagnostic to remediation, in one pipeline. A confirmed read routes to a certified remediator or to a laboratory, and the outcome routes back. The commercial loop and the learning loop are the same loop, which is why the dataset compounds rather than ages. Proprietary scans, fused with open data. Scans are augmented with 5 curated open-data layers, moving the product from reading the wall in front of you toward predictive property-level mould-risk scoring for the ones you have not visited. An environmental data platform underneath. Weather, humidity history, flood severity and air quality already run in production behind the live conditions on this site. It is the same platform an insurer or a strata portfolio would license, and it is built and running. The entry barrier works from the other direction too. A $19.49/mo subscription and a 60-second scan sit against $300+ lab tests, so the cheapest way into the category is also the fastest way to feed it. THE DATA FLYWHEEL ### What a competitor would have to rebuild: years of data and partners. Inputs at the edges feed the lake; the wheel turns as each stage feeds the next. - Weather and climate. Live Copernicus ERA5-Land history and NOAA GFS forecasts, collected for Australia today. - Flood and air quality. Live GloFAS river-flood forecasts and CAMS air quality, in the same lake. - Building-remediation images. Partners Photographs from a remediation network with 1,500+ verified remediation projects behind it. - Lab results from partners. Collecting Lab results collated through approved remediators, cleaners and lab partners; collation is under way and becomes the highest-grade ground truth. - Specialist assessments. Beta Room-by-room reports from the Mould Assessment app, saved as structured data. AI has made software cheap to write, so the app alone is easy to copy. The economic model that brings trades, labs and remediators onto one platform, the partnerships behind it, and the weather, environmental, remediation-image and partner lab data they accumulate are much harder to copy. PATENTABLE AI TECHNOLOGY ### Public science. Proprietary application. Every technique used to build our convolutional neural networks for image analysis is publicly available. The specific application is not. Combining them into an instrument that reads real walls in real light takes domain expertise across AI and machine learning algorithms and architectures that is very hard to reproduce, even with AI-assisted coding. The image-processing pipeline and the CPU- and GPU-efficient approaches to model training and inference are unique to Mould Detect. A draft patent specification covering them has been prepared for our attorneys, with protection sought at several levels and drafted against current United Kingdom and Australian subject-matter law. AI Vision has been reviewed by two patent lawyers and confirmed as worthy of a patent application. THE MOULD ATLAS · AND IOT ### A prediction engine, fed from the sky and from inside buildings. Sources on the left feed the lake; the lake feeds the Mould Atlas. Filled phases run today; outlined phases are the roadmap. - Independent environmental feeds. Live ERA5-Land climate history, NOAA GFS forecasts, GloFAS river flood and CAMS air quality, collected for Australia today. - Readings from inside buildings. Prototype Sensor boards post to the same lake on a sensor-agnostic schema: a new sensor adds a key, not a new pipeline. - Scans, assessments and lab results. Collecting AI Vision scans, Mould Assessment reports and partner lab results become the labels the Atlas learns from. - Phases 0 and 1 run today. Live The data lake, and a published VTT mould-growth model per location served as /v1/mould-index. It is driven by outdoor conditions and has not yet been validated against indoor damage. - Phases 2 to 4 are the roadmap. Roadmap Learn the outdoor-to-indoor transfer by building type, train on verified outcomes, then predict risk by building type, postcode, climate zone and season. IoT hardware is outside the MVP and year-1 revenue: a prototype board feeds the lake today, and a sensing diffuser is planned as a 2027 prototype. [See the hardware and its phases.](https://moulddetect.ai/iot/) COLLABORATIVE BUSINESS MODEL ### An ecosystem, not only an app. Mould Detect is not just an AI mould detection app. It is an ecosystem: environmental health specialists like [Natures Defence](https://naturesdefence.co/), building remediators like [Keystone Building Solutions](https://keystonebs.com.au/), ventilation specialists like [Riteflo](https://riteflo.com.au/), mould remediators, laboratories and environmental scientists like Civil and Domestic Environmental Sciences, occupational hygienists, and cleaning and restoration experts, many certified through the [Australian Cleaning and Restoration Academy](https://cleaningrestorationacademy.com.au/). THE MODEL ### Four ways the same scan earns. One product, one ladder, two audiences. A household pays for peace of mind and a trade pays for a report it can invoice against, and both climb the same subscription rungs as their needs grow. The public pricing page is the single source of tier truth. Subscriptions, $19.49 to $399+ per month. Homes start at $19.49/mo. Specialists climb through Business and Professional to an Enterprise tier at $399+ that carries white label, API access and insurer-ready data. The free tier is a limited monthly allowance, set at launch. Marketplace commissions. Confirmed cases route to certified remediators and to product suppliers, and the routing itself is a revenue line rather than a courtesy. Laboratory testing kits. Where a read calls for confirmation, the platform sells the kit that provides it through laboratory partners, so the ecosystem is monetised rather than lost at the referral. API access and data licensing. The environmental platform and the scan corpus are sold as data products to labs, insurers and strata portfolios, which is the stream that carries the moat into revenue. [See the full tier ladder, feature by feature.](https://moulddetect.ai/pricing/) The regulatory tailwind. Regulation is moving toward us. Awaab's Law obliges landlords in the United Kingdom to act on damp and mould within fixed timeframes, and the National Construction Code's 2025 condensation management provisions push the same way in Australia. Both create demand for assessments that are timestamped, evidenced and easy to produce on request, which is the artefact this product makes by default. AWAAB'S LAW (UK) IN FORCE 27 OCT 2025 · NCC 2025 CONDENSATION MANAGEMENT (AU) GO-TO-MARKET · IN SUMMARY ### Trades first, homes through the free tier, then enterprise. The sequence, at the level of detail that belongs on a public page. The deck carries the rest. - #### The wedge: property-servicing trades. Cleaners, remediators and plumbers who scan across many properties are the primary revenue, reached professionally and through their communities. - #### Homes as the front door. A free tier and a $19.49 plan bring households in and route them to professionals and labs when they need one. - #### Then enterprise, and new markets. Strata and enterprise anchors follow the proven wedge. Australia proves the approach and the technology first; the UK, where Awaab's Law applies, is next, then Canada and the US at Series A. WHERE WE LAUNCH ### Launching in Australia, then the UK, Canada and the US. In order: the markets we plan to enter, and why. The chips say which is live today. - Now #### Australia: prove the approach and the technology. Australia is our testing ground. One in two homes have damp or mould, our founders' trade network is here, and our environmental data covers Australia. - Planned #### The United Kingdom, next. The UK puts a legal clock on mould: Awaab's Law, in force since 27 October 2025. Fast triage and a dated record help landlords act in time. Entry is gated on revenue. - Planned #### Canada and the US, at Series A. Canada and the US follow, after the UK. Where claims and property deals rely on accredited lab results, as in the US, a partner lab supplies the result and Mould Detect supplies the triage and referral. ADJACENT MARKETS ### The same pipeline reads more than mould. Where the computer-vision pipeline transfers, nearest first. None of these is committed. - #### Building defects. Cracks and water damage that remedial builders inspect by eye today; the pipeline that classifies mould can flag them. - #### Crop disease and food security. Early detection in agriculture uses the same convolutional and segmentation techniques, and research groups are already working that angle. - #### Digital pathology and radiology. Counting cells and flagging findings use versions of the same approach, though neither is a market Mould Detect is building toward now. THE TEAM ### One founding team, three disciplines. One founding team of seven, led by three co-founders. Mould Detect was founded by a remedial builder with more than twenty years on site, an environmental health specialist in natural remediation, and a technology founder who has built enterprise data platforms before. Domain, science and engineering, one each, with the remediation network that supplies the dataset already in the room. [Read the founding team in full.](https://moulddetect.ai/about/) Sources - Mould Detect concise pitch deck · 14 October 2026 · market sizing, platform statuses, team - Mould Detect · Ventures Australia 2026 deck, appendix A12 and A12b · market sequence: Australia, then the UK, then Canada and the US - Mould Detect business plan v3.1 · 14 October 2026 · claims precedent, Mould Atlas phases, IoT, adjacencies - [GOV.UK · Awaab's Law guidance for social landlords](https://www.gov.uk/government/publications/awaabs-law-guidance-for-social-landlords/awaabs-law-guidance-for-social-landlords-timeframes-for-repairs-in-the-social-rented-sector) - [NHS · Sick building syndrome](https://www.nhs.uk/conditions/sick-building-syndrome/) Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ON REQUEST ### The detailed numbers go to investors directly. Financial projections, unit economics and the details of our seed round are shared directly with investors. This page stops at the market sizing, the year-3 forecast and the public pricing. The rest comes with a conversation. The pitch deck. The current version, with traction and the roadmap. Financial projections. The revenue path year by year, beyond the year-3 forecast on this page. Unit economics. Acquisition cost, lifetime value, retention and margin, with the assumptions behind each. The seed round. The amount, the use of funds, the terms and the runway. Time with the founders. A call or a meeting with the three co-founders. Ask for any of them below. INVESTOR ENQUIRIES ### Ask for what you need to decide. Tell us who you are and what you would like to see. We reply by email. --- # Mould Detect for IoT and Embedded Devices | Sensor Boards and the Data Lake URL: https://moulddetect.ai/iot/ FOR IOT AND EMBEDDED DEVICES ## Sensor readings that join the data lake. A photo tells you what is on the wall; readings from inside the building tell you why. Mould Detect's prototype board already posts into the same data lake as every scan and weather feed, on a schema where a new sensor adds a key, never a new pipeline. THE HARDWARE ### A board running today, and a diffuser in design. Left, a recording of the test board feeding the live dashboard. Right, the concept design of the sensing diffuser, numbered part by part. The ESP8266 test board, on a battery pack, feeding the live dashboard on a laptop Prototype An ESP8266 Witty board posts batched HTTPS readings about every 30 seconds in development. Its one light sensor is there to prove the pipeline; the physics comes next. Planned 2027 The sensing diffuser, from the Ventures Australia 2026 deck: ultrasonic diffusion plus a sensor board for temperature, humidity, moisture, formaldehyde, carbon dioxide, UV and VOCs. A concept in R&D, planned as a 2027 prototype. Concept design by Rozario Chivers, Vladimir Lekic and Lucas Chivers. BUILD PHASES ### What each phase adds, and where it stands. Read top to bottom. Every phase posts into the same schema, so nothing downstream changes as sensors are added. Mould Detect IoT build phases and status Phase Status What it adds 1 to 2b Done A local prototype, cloud ingest into the data lake, and a hosted read-only dashboard. 3 Next build Temperature, relative humidity, dew point and pressure (SHT31 or BME280), and a calibrated light sensor. 4 Roadmap On and beside the diffuser: VOC (SGP40), carbon dioxide (SCD40), formaldehyde, UV, surface moisture, water depth and water leak, plus the diffuser's own on/off and duty state. Particulate sensing is not planned: outdoor particulates already come from CAMS. Water depth and leak sensing matter after floods and plumbing failures, which raise indoor humidity. MOULD DETECT BUSINESS PLAN v3.1 · 14 OCT 2026 THE PIPELINE ### Cheap to run, and the same schema for every sensor. Three design decisions an engineer will ask about first. - #### Device, time, key, value. Readings are stored as four fields, so a new sensor adds a key rather than a table, a stream or a service. - #### Serverless ingest. Device to a Lambda function URL, into the S3 lake partitioned by device and day, rolled up and served through a read API to the dashboard. - #### About 2.3 US cents per device a month. Modelled running cost at the 15-minute production cadence. Idle costs nothing, because nothing is always on. WHERE READINGS GO ### Indoor readings teach the Atlas what outdoor data cannot. Sources on the left feed the lake; the lake feeds the Mould Atlas. Sensor boards are how phase 2, indoor calibration, will be learned. - Independent environmental feeds. Live ERA5-Land climate history, NOAA GFS forecasts, GloFAS river flood and CAMS air quality, collected for Australia today. - Readings from inside buildings. Prototype Sensor boards post to the same lake on a sensor-agnostic schema: a new sensor adds a key, not a new pipeline. - Scans, assessments and lab results. Collecting AI Vision scans, Mould Assessment reports and partner lab results become the labels the Atlas learns from. - Phases 0 and 1 run today. Live The data lake, and a published VTT mould-growth model per location served as /v1/mould-index. It is driven by outdoor conditions and has not yet been validated against indoor damage. - Phases 2 to 4 are the roadmap. Roadmap Learn the outdoor-to-indoor transfer by building type, train on verified outcomes, then predict risk by building type, postcode, climate zone and season. FROM READING TO RESPONSE ### A reading that crosses a threshold can act. Read top to bottom: data sources, a threshold agreed in advance, and the responses that fire. - Oracles. Live Independent data no party to the contract controls: weather and flood feeds, in-building sensors, photo scans, specialist assessments and lab results. - A threshold agreed in advance. Illustrative For example relative humidity above X% for N days, a GloFAS flood exceedance, or a condensation risk from the dew-point spread. - A. Environmental control. Phase 4 Ventilation, a dehumidifier or the diffuser switched on when a trigger fires. - B. Alerts. Prototype IoT alerting to landlords, facilities and maintenance teams, early. - C. Remediation booking. Partners A lab test or an approved remediator, booked through the supplier list. - D. Data and API feed. Live The same readings delivered to partners, including insurers, into their own systems. The technical pattern only. No insurer pilot or contract exists today. Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ### Build sensors with us. If you design boards, firmware or building sensors, tell us what you measure. The schema is ready for a new key. [Email us](https://moulddetect.aimailto:hello@moulddetect.ai?subject=IoT%20and%20embedded)[See the live dashboard](https://iot.moulddetect.ai/) --- # Learn About Mould | Mould Detect URL: https://moulddetect.ai/learn/ MOULD EDU ## Know the mycological terrain, we provide the map. Practical, Human health and building health guidance, understand the mould risk factors, how to clean effectively, how to prevent mould coming back, and the natural approaches proven to work by experienced professionals in the field. HOW WE HELP Mould Detect gives you tools, information and support for confident decisions. If you suspect a mould issue, use our apps to understand and document what you are seeing, and always seek professional advice. Where possible, confirm with laboratory testing to fully understand your situation. [Join the early access](https://moulddetect.ai/#early-access) THE SPECIES ### Know what you are looking at. Thousands of mould species exist, but a handful account for most indoor problems. Knowing what you are looking at helps you judge urgency. #### Stachybotrys chartarum, the black mould Slimy, dark green-black patches on chronically wet plasterboard, timber or paper. It produces mycotoxins linked to respiratory irritation, so treat any suspected patch with caution and avoid dry brushing. #### Aspergillus Very common indoors: powdery green, yellow or brown colonies. Most species are harmless to healthy people, but they can cause serious infection in those with weakened immune systems and can aggravate asthma. #### Penicillium Blue-green, velvety growth on water-damaged materials, fabrics and food. It spreads quickly and is a frequent allergy trigger, bringing sneezing, itchy eyes and sinus congestion. #### Cladosporium Olive-green to brown-black, and unusual in that it thrives in both warm and cool areas: window frames, bathroom seals, air-conditioning vents. A common asthma and hay fever trigger. #### Alternaria Dark grey-brown with a wool-like texture, common around showers, tubs and leaky sinks. One of the strongest allergenic moulds, and often the reason behind a persistent indoor hay fever. SAFE CLEANING ### Clean it safely, or know when not to. Small, surface-level patches, under about half a square metre, can usually be handled yourself. Anything larger, recurring, or inside wall cavities needs a professional. #### Confirm it is mould Mould smears when rubbed with a damp cloth and usually smells musty; dirt does not. A Mould Detect scan gives you a fast first opinion and a record of the area over time. #### Protect yourself first Wear gloves, a P2/N95 mask and eye protection. Open windows, close internal doors, and never dry-brush or vacuum mould, because that launches spores into the air. #### Clean, do not just bleach Bleach whitens but often leaves the roots in porous surfaces. Wash with warm soapy water or a vinegar solution, dry thoroughly, and keep watching the spot. Discard porous items such as carpet and ceiling tiles that stay mouldy after cleaning. #### Know when to stop Patches larger than about 0.5 m2, mould returning after cleaning, growth related to sewage, or anyone in the home with a respiratory condition: at that point, bring in a professional. [ WHITE-LABEL REPORTS BUSINESS TIER AND UP . See pricing. ](https://moulddetect.ai/pricing/) PREVENTION ### Take away the moisture, take away the mould. Mould needs three things: humidity above about 60%, something to eat, which ordinary household dust supplies, and around 48 hours. Remove the moisture and you remove the mould's future. #### Control humidity Keep indoor humidity below 60%, and ideally between 30 and 50%. Use exhaust fans when cooking and showering, and a dehumidifier in damp rooms or damp seasons. #### Move the air Open opposing windows daily for cross-flow, leave wardrobe and cupboard doors ajar in damp weather, and keep furniture a hand-width off external walls. #### Fix water fast Leaks, rising damp and condensation feed mould within 24 to 48 hours. Dry wet areas immediately and repair the source, not just the stain. #### Watch the cold spots Window frames, south-facing walls, built-in wardrobes and bathroom ceilings are condensation magnets. Check them monthly, and scan anything suspicious. NATURAL SOLUTIONS ### Effective without the harsh fumes. For routine cleaning and prevention, several natural options are effective and avoid harsh fumes. It is the approach our partners at Nature's Defence specialise in. #### White vinegar Undiluted white vinegar kills most household mould species and penetrates porous surfaces better than bleach. Spray it on, leave it for an hour, then wipe and dry. #### Clove oil A few drops, about a quarter teaspoon per litre of water, inhibits mould regrowth on hard surfaces. Use it sparingly, because it is potent, and keep it away from pets. #### Tea tree oil Two teaspoons per two cups of water makes an effective anti-fungal spray for bathrooms and window frames. There is no need to rinse; the residue keeps working. #### Bicarb and sunlight Bicarbonate of soda scrubs and deodorises, and direct sunlight is a free anti-mould treatment. Air bedding, rugs and cushions outside regularly. WATCH AND LEARN ### Video guides from our partners. - [ #### Ep 01 The Mystery Of Mold - What Every Homeowner Must Know NATURE'S DEFENCE ](https://www.youtube.com/watch?v=uhxh7NERjZ8) - [ #### Safe Mould Removal: Natural Solution and Holistic Approach NATURE'S DEFENCE ](https://www.youtube.com/watch?v=gRCFC3GQOko) - [ #### Discovering Mycotoxins: What You Need to Know NATURE'S DEFENCE ](https://www.youtube.com/watch?v=OO6w131ghC8) - [ #### Mould Reset: Kill Existing Mould, Fix the Fault, Maintain Your Home NATURE'S DEFENCE ](https://www.youtube.com/watch?v=5hNOGBkbqrI) - [ #### Mould Stains NATURE'S DEFENCE ](https://www.youtube.com/watch?v=Xm1GuXdyKok) - [ #### About Our Video Training Series: How to Become the Best Carpet and Upholstery Cleaning Technician AUSTRALIAN CLEANING & RESTORATION ACADEMY ](https://www.youtube.com/watch?v=aZ2bOh4OM9o) - [ #### About Our Video Training Series on Restoration AUSTRALIAN CLEANING & RESTORATION ACADEMY ](https://www.youtube.com/watch?v=eQJvK_t7Z6o) - [ #### About Our Training Video on Work Place Health and Safety AUSTRALIAN CLEANING & RESTORATION ACADEMY ](https://www.youtube.com/watch?v=IjL9MSLOI1k) CERTIFICATIONS ### Learn at your own pace, earn the credential. Structured courses and certifications in mould remediation are coming to Mould Edu. Learn at your own pace and earn recognised credentials. COMING SOON [ EDUCATION AND CERTIFICATION BUSINESS TIER AND UP . See pricing. ](https://moulddetect.ai/pricing/) MISS MOULD AI ### Ask the mycology expert. Any question, plain answers. - YOU Is the black patch behind my wardrobe dangerous? - MISS MOULD Possibly. Dark growth on a cold external wall behind furniture is a classic condensation pattern. Photograph it in good light, run a scan, and do not dry-brush it. If it is larger than your hand or comes back after cleaning, bring in a professional. Miss Mould AI lives in the Mould Detect app. [Join the early access](https://moulddetect.ai/#early-access)[ MISS MOULD AI PROFESSIONAL TIER AND UP . See pricing. ](https://moulddetect.ai/pricing/) --- # Partner with Mould Detect | Trades, Inspectors, Hygienists and Labs URL: https://moulddetect.ai/partners/ FOR PARTNERS ## Build the mould dataset with us. The people who walk into damp rooms every week see more mould than any camera roll. Mould Detect partners with the trades, inspectors, hygienists and laboratories who capture those images and readings, routes work back to them, and opens its APIs to the products they build. We are launching in Australia to prove the approach and the technology, with the UK, Canada and the US to follow. WHO WE ARE LOOKING FOR ### If you see damp at work, your photographs are worth more. Find your trade. Each card says what you would capture, and what it earns you. - #### Contract cleaners Add a mould scan to every clean. It sets your service apart, justifies a premium, and keeps clients longer. You capture images and conditions room by room. - #### Plumbers Leaks are where mould starts. A scan at the source documents it, and the inspection can route on to a remediation partner. - #### Shower fitters and bathroom installers Bathrooms show damp first. A dated scan before and after a job records the condition you found and the condition you left. - #### Building inspectors An on-site risk read with the local weather, humidity and flood context attached, captured as structured, location-tagged evidence. - #### Mould inspectors and remediators Scan every property before and after treatment for documented evidence, and assess on site before you quote, which reduces callbacks. - #### Occupational hygienists and laboratories Ecosystem partners, never competitors. Likely-positive scans route to lab testing, and lab-verified results become the highest-grade labels in the dataset. [OH AND LABS ](https://moulddetect.ai/for-specialists/#oh-and-labs) WHO WE WORK WITH TODAY ### Labs, cloud, trades and research partners. One card per partner and what they do with us. - #### CDES Lab testing and species images - #### Cloud Assembly AWS Partner, built the Vision API - #### AWS Startup Program - #### Spotless Interior Cleaning Tests our prototype and assessment app; offering training images - #### Riteflo Ventilation and Damp Rods (co-founder's company); staff testing and using the assessment app - #### ACRA Training and cleaner certification - #### Nature's Defence Natural treatment products (co-founder's company) - #### Brock University Research partnership WHAT YOUR CAPTURE BUILDS ### Images and readings that make every model better. Inputs at the edges feed the data lake in the centre; the wheel turns as each stage feeds the next. Partner images and lab results are the inputs nobody can buy. - Weather and climate. Live Copernicus ERA5-Land history and NOAA GFS forecasts, collected for Australia today. - Flood and air quality. Live GloFAS river-flood forecasts and CAMS air quality, in the same lake. - Building-remediation images. Partners Photographs from a remediation network with 1,500+ verified remediation projects behind it. - Lab results from partners. Collecting Lab results collated through approved remediators, cleaners and lab partners; collation is under way and becomes the highest-grade ground truth. - Specialist assessments. Beta Room-by-room reports from the Mould Assessment app, saved as structured data. ENVIRONMENTAL DATA IN THE APP ### See the conditions behind the damp before you decide what to do. The cards give the conditions at a property now; the chart shows how long humidity has sat above the 70% damp line, which says more than any single reading. Every reading shown here is a sample. The app's environment view for specialists: weather, humidity and air quality Why it helps your business - Explain the cause. Show a client how long humidity sat above the damp line, so they see why the mould grew. - Plan the work. The Ventilation Planner shows when opening the windows helps and when to keep them shut. - Keep it together. The scan result and local conditions sit side by side in one app. - Weather Light rain - Temperature 18 °C - Humidity 82% · Damp - Air quality PM2.5 · Good Sample readings Air quality: Generated using Copernicus Atmosphere Monitoring Service information 2026 Humidity history Sample humidity history: 12 daily readings from 62% to 82%, at or above the 70% damp line on 8 of them. A photo shows the mould. The weather, temperature, humidity history and air quality for the same location show what is feeding it, so a professional can judge whether it will come back after the clean. The data comes from the platform behind our public API and covers Australia. YOUR EVIDENCE ### A report you can hand to a client, under your own logo. The clip shows a scan becoming a report: photos, heat maps, environmental readings and a next best action. A scan result turned into a shareable report in the app White-label reports start on the Business plan. [Compare the plans.](https://moulddetect.ai/pricing/) AGENT FIRST ### Add mould intelligence to your own product. Four building blocks, each with its status today. Partners can combine any of them with what they already sell. - Built #### Mould detection. The AI Vision model that reads a photo of a suspect surface, available to partners with developer key access. [AI VISION ](https://moulddetect.ai/ai-vision/) - Live #### Environmental data. Mould growth index, river flood forecasts, air quality and site layers for any Australian point, over a documented API. [DEVELOPERS ](https://moulddetect.ai/developers/) - Live #### Weather data. Daily forecasts, current conditions and ERA5-Land weather history, with units and sources in every response. [API REFERENCE ](https://environment.moulddetect.ai/docs/) - Beta #### AI agents. An MCP server and terminal agent that put the same data inside Claude, Codex, Cursor and your own agents. Agent first: built for software that reads, as well as people. [AGENTS AND MCP ](https://moulddetect.ai/agents/) WHERE WE LAUNCH ### Launching in Australia, then the UK, Canada and the US. In order: the markets we plan to enter, and why. The chips say which is live today. - Now #### Australia: prove the approach and the technology. Australia is our testing ground. One in two homes have damp or mould, our founders' trade network is here, and our environmental data covers Australia. - Planned #### The United Kingdom, next. The UK puts a legal clock on mould: Awaab's Law, in force since 27 October 2025. Fast triage and a dated record help landlords act in time. Entry is gated on revenue. - Planned #### Canada and the US, at Series A. Canada and the US follow, after the UK. Where claims and property deals rely on accredited lab results, as in the US, a partner lab supplies the result and Mould Detect supplies the triage and referral. Other doors People and homes - [ Homes & renters ](https://moulddetect.ai/homes/) - [ Specialists & pros ](https://moulddetect.ai/for-specialists/) Organisations - [ Insurers ](https://moulddetect.ai/insurers/) - [ Investors ](https://moulddetect.ai/investors/) - [ Partners ](https://moulddetect.ai/partners/) Builders of technology - [ Developers ](https://moulddetect.ai/developers/) - [ AI agents ](https://moulddetect.ai/agents/) - [ IoT & devices ](https://moulddetect.ai/iot/) ### Become a partner. Tell us your trade, where you work and what you would like to build. We will reply with how the partnership works for you. [Email us](https://moulddetect.aimailto:hello@moulddetect.ai?subject=Partner%20enquiry)[For specialists](https://moulddetect.ai/for-specialists/) HELLO@MOULDDETECT.AI --- # Pricing | Mould Detect URL: https://moulddetect.ai/pricing/ PRICING ## Free to try. Specialists scale up. No trial period, no card required: the Free tier stays free for as long as you use it. Specialists who need more scans, more properties and reports built to hand to a client can step up the ladder whenever they are ready. THE PATHWAY ### Five tiers, one ladder. Free stays free. Specialists step up for more scans, more properties and reports built to hand to a client. CANCEL ANYTIME · NO LOCK-IN · ALL PRICES AUD Mould Detect pricing tiers and the features each one includes. FOR YOUR HOME FOR SPECIALISTS Plan Free Renters and occasional users. [Get early access](https://moulddetect.ai/#early-access) No card required $0 /month $0 /month Essential Homeowners. [Get early access](https://moulddetect.ai/#early-access) $19.49 /month $199.99 /year Save ~14% Most popular Business Small and medium firms: cleaners, trades, sole traders. [Get early access](https://moulddetect.ai/#early-access) $49 /month $489.99 /year 2 months free Professional Remediation firms, franchises and inspection practices. [Get early access](https://moulddetect.ai/#early-access) $199 /month $199 /month Monthly only Enterprise Labs, insurers and mycological partners. [Get early access](https://moulddetect.ai/#early-access) From $399+ /month $399+ /month Monthly only Scanning & AI AI mode Hybrid AI: cloud at launch, edge phased in Cloud AI Cloud AI Cloud AI Cloud AI Scans per month 10 100 300 1,000 Unlimited Properties 1 2 10 20 Unlimited Confidence score Score only Full Full Full Full Risk score Not included Full Full Full Full Miss Mould AI Limited access Limited access Limited access Included Included Records & reporting Health history On device only Included Included Included Included Saved data On device only Cloud backup Cloud backup Cloud backup Cloud backup Saved reports Not included Included Included Included Included White-label PDF reports Not included Not included Included Included Included Data & integrations Weather & environmental mould-risk data Limited access Limited access Limited access Included Included API access Not included Not included Not included Not included Included Insurer-ready data Not included Not included Not included Not included Included Growing your practice Education, training & certification Not included Not included Included Included Included Free Renters and occasional users. [Get early access](https://moulddetect.ai/#early-access) No card required $0 /month $0 /month Start with: - AI mode Hybrid AI: cloud at launch, edge phased in - Scans per month 10 - Properties 1 - Confidence score Score only - Miss Mould AI Limited access - Health history On device only - Saved data On device only - Weather & environmental mould-risk data Limited access Essential Homeowners. [Get early access](https://moulddetect.ai/#early-access) $19.49 /month $199.99 /year Save ~14% Everything in Free, plus: - AI mode Cloud AI - Scans per month 100 - Properties 2 - Confidence score Full - Risk score Full - Health history ✓ Included - Saved data Cloud backup - Saved reports ✓ Included Most popular Business Small and medium firms: cleaners, trades, sole traders. [Get early access](https://moulddetect.ai/#early-access) $49 /month $489.99 /year 2 months free Everything in Essential, plus: - Scans per month 300 - Properties 10 - White-label PDF reports ✓ Included - Education, training & certification ✓ Included Professional Remediation firms, franchises and inspection practices. [Get early access](https://moulddetect.ai/#early-access) $199 /month $199 /month Monthly only Everything in Business, plus: - Scans per month 1,000 - Properties 20 - Miss Mould AI ✓ Included - Weather & environmental mould-risk data ✓ Included Enterprise Labs, insurers and mycological partners. [Get early access](https://moulddetect.ai/#early-access) From $399+ /month $399+ /month Monthly only Everything in Professional, plus: - Scans per month Unlimited - Properties Unlimited - API access ✓ Included - Insurer-ready data ✓ Included THE SME ACCELERATOR ### More than software. A practice, accelerated. A subscription runs your inspections, sharpens them with live environmental insight, connects you to experts and laboratories, and certifies you as you grow. Your inspections, managed. Every scan is recorded, location-tagged and report-ready, so the paperwork side of an inspection runs itself. Insight on every job. Weather, humidity, flood-risk and air-quality context turns a reading on a wall into advice a customer can act on. A network, not just a tool. Access to other professionals, experts and laboratories: escalate, refer and confirm without leaving the platform. Education that scales you. Training and certification from Business tier up, so the service you sell grows as your subscription does. WHAT UNLOCKS WHEN - White-label PDFs Business and up - API access + insurer-ready data Enterprise PRICING FAQ ### Straight answers on cost. #### What does the Free tier actually include? A monthly allowance of 10 scans for one property, with a basic confidence score only. Free uses hybrid AI: cloud at launch, with on-device edge AI phased in by demand and cost; every paid tier runs on cloud AI. It stays free for as long as you use it, with no trial period and no card required. #### Can I upgrade or downgrade at any time? Yes. Move between tiers whenever your needs change. An upgrade or a downgrade takes effect from your next billing cycle, so you are never paying for a tier you have already moved off. #### What is the real difference between Free and Essential? Free carries a basic confidence score, a monthly scan allowance and scan history kept on your device only. Essential lifts the allowance to 100 scans a month and adds the full risk score, health history, saved reports and cloud backup for two properties, from $19.49 a month. #### Which tier fits a solo cleaner or trade? Business, most often: 300 scans a month across ten properties, plus white-label PDF reports you can hand straight to a client under your own name. #### I run a laboratory or work in occupational hygiene. Where do I fit? Enterprise: unlimited scans and properties, API access, insurer-ready data exports and education, training and certification support. The for specialists page has the fuller picture. #### Are these prices final? Prices are per the July 2026 pricing model and quoted in Australian dollars. They are subject to change before public launch. PRICES PER THE JULY 2026 MODEL · AUD · SUBJECT TO CHANGE BEFORE LAUNCH ### Start free. Step up when you are ready. Early access is opening now, at every tier from Free through to Enterprise. [Join the early access](https://moulddetect.ai/#early-access) --- # Privacy | Mould Detect URL: https://moulddetect.ai/privacy/ PRIVACY AND DATA ## Your data, handled with care. What we collect across our website, data APIs, Mycelium Network and iOS app, why we collect it, how long we keep it and what you can ask of us. Last updated 31 October 2026. ABOUT THIS POLICY ### Who we are and what this covers. This policy explains how MouldDetect Technologies Pty Ltd (ABN 80 696 130 784) handles personal information. “We”, “us” and “Mould Detect” mean that company. Last updated 31 October 2026. It covers four services: - Our public website, moulddetect.ai - Our data APIs, data.moulddetect.ai and environment.moulddetect.ai - The Mycelium Network: our MCP server and the Mycelium command-line tool (CLI) - The Mould Detect iOS app and the services it calls Mould Detect is for adults. You need to be 18 or older to use the app, create an account or hold an API key. OUR APPROACH ### Only what a feature needs. Personal information is information or an opinion about you, or about someone who could reasonably be identified. We follow the Australian Privacy Principles in the Privacy Act 1988 (Cth). We collect only what a feature needs, we keep it in Australia wherever we can, and we tell you below where we cannot. We do not sell personal information and we do not use it for advertising. THE WEBSITE ### No cookies, and two forms. - Cookies and storage. The website sets no cookies and runs no analytics or advertising tools. Your browser keeps one preference, your light or dark theme choice, in its own local storage on your device. We never receive it. - Fonts are served from our own site, so loading a page sends nothing to a font provider. - Videos on /learn/. The video thumbnails are loaded from YouTube, so your browser asks Google for them and Google receives your internet address and browser details. Playing a video happens on YouTube under Google’s terms. - Early-access waitlist. We keep your email address, when you joined and confirmed, and a salted one-way hash of your internet address, used only to limit abuse. Joining is double opt-in: you are on the list only after you click the link in our confirmation email; if you never confirm, we delete your sign-up after 30 days. Every email we send has a withdraw link. If you withdraw, we delete your sign-up 30 days later. All we keep after that is a scrambled (hashed) form of your email address on a do-not-email list, so we never email you again unless you sign up afresh. Otherwise we keep your sign-up until you ask us to delete it. - Enquiry forms. We keep what you enter (your name, email, organisation if given, the options you tick and your message), the page and time, and a salted one-way hash of your internet address. The enquiry is stored in AWS Sydney and emailed to our team inbox, hello@moulddetect.ai, through Amazon’s email service. We use it only to reply and do not add you to a mailing list. We keep it until you ask us to delete it. - Service logs. The service behind the waitlist and enquiry forms keeps technical logs for 30 days. THE DATA APIS ### We do not log the places you look up. data.moulddetect.ai and environment.moulddetect.ai return weather, humidity, flood and air-quality data for places you ask about. - We do not log the coordinates you query. - If we issue you an API key, we store only a hash of the key, never the key itself, together with the name of the app it was issued for and daily request counts used to apply your quota. THE MYCELIUM NETWORK ### Tools for AI assistants, kept briefly. The Mycelium Network gives AI assistants and the Mycelium CLI access to our mould-risk, weather, flood, air-quality and photo tools. - What we keep. A hash of your key and your account identifier (from Amazon Cognito). Usage records, used for your daily allowance, are kept for 35 days. Service logs are kept for 14 days. - Photos sent through the photo tool go to the same vision service as the app and are not kept. - The CLI and your AI provider. The CLI sends your prompts to the AI provider you choose (Anthropic, OpenAI or Google), using your own account with that provider. That provider handles your prompts under its own terms and privacy policy, not ours. - AI assistants. If you connect our MCP server to an AI assistant, that assistant’s provider handles your conversation under its own terms. We receive only the tool requests it sends us. THE IOS APP ### Your photo is checked, then let go. - Your account. Sign-in runs on Amazon Cognito in AWS Sydney. We hold your email address and an account identifier. Your password is held by Cognito, not by us. - Checking a photo. Before a photo leaves your phone, the app removes its location data and camera details. The photo is then sent to our vision service in AWS Sydney without your account identifier, name or email. The service analyses it, returns the result and heat map, and does not keep the photo. When you are offline, the app can analyse the photo entirely on your phone. - Your scan library. Each scan keeps the photo, result, heat map and any details you add (property, room, surface, smell, lighting, notes). It can also record approximate location (about 1 km), the phone’s angle, air pressure, light level, and your phone model and iOS version. This stays on your phone unless you back it up. - Cloud backups (optional, paid plans). When you choose to back up, copies of your scans, including their photos, notes and approximate location, are stored in AWS Sydney in a private area linked to your account, so that only you can restore them. Backups are linked to you; they are not anonymised. - Location. The app asks for approximate location only, and only for a feature that needs it, such as filling in an address, finding nearby specialists or recording where a scan was taken. Turning coordinates into an address uses Apple’s map service. - Phone permissions. Camera, photo library, location, motion and microphone are each asked for separately. You can refuse or turn off any of them in Settings; only the feature that needs it stops working. - Contacting a specialist. The app opens your own email app with a message you write. It goes directly from you to the specialist; we do not receive it. - Subscriptions. Apple handles payment. We receive your subscription status, not your payment details. AI TRAINING AND ANALYSIS ### Your photos do not train our models today. We do not currently use your photos to train AI models. If that changes, we will ask for your opt-in consent first, and any photo used will have its location data, camera details (EXIF) and link to your account removed. Mould results and heat maps come from our own AI models without a person reviewing them. They support your decisions; they are not a diagnosis, a laboratory test or a building inspection. HEALTH DISCLAIMER ### Decision support, not a diagnosis. Mould Detect provides AI-based identification for decision support and educational purposes only. This application is NOT a substitute for professional health advice, medical diagnosis or certified building inspection services. AI predictions may carry margins of error arising from lighting, resolution and surface complexity. Always consult a certified health care professional, toxicologist or environmental specialist for health concerns. WHO ELSE RECEIVES INFORMATION ### Our providers, and the ones you choose. We use service providers that handle information for us under contract and only on our instructions, chiefly Amazon Web Services in Sydney (hosting, sign-in, storage, backups, photo analysis and email). Other organisations receive information because of how you use our services: - Apple: App Store, payments and map lookups, under Apple’s privacy policy - Google: YouTube thumbnails on /learn/ - The AI provider you choose for the Mycelium CLI, or the provider of the AI assistant you connect - Specialists you email yourself from the app We do not otherwise disclose personal information unless the law requires it. INFORMATION THAT LEAVES AUSTRALIA ### Sydney by default. We store and process your information in AWS Sydney. Information goes overseas only in these cases: - Google (United States and other countries), when your browser loads YouTube thumbnails - The AI provider you choose for the Mycelium CLI (Anthropic, OpenAI or Google), which may process your prompts in the United States or elsewhere - Apple (United States and other countries), for the App Store, payments and map lookups SECURITY ### How we protect it. Information travels encrypted (HTTPS). API and Mycelium keys are stored only as hashes. Backups sit in private storage that blocks public access, and each account can reach only its own files. Photos sent for analysis carry no account details and are not stored. If you think someone has used your account or key, tell us at privacy@moulddetect.ai. HOW LONG WE KEEP IT ### A stated time, or until you ask. - Photos sent for analysis: not kept. - Scans on your phone: until you delete them. - Cloud backups: until you close your account or ask us to delete them. Deleting a scan on your phone does not yet remove its backup copy; ask us and we will. - Mycelium usage records: 35 days. - Mycelium service logs: 14 days. - Waitlist sign-ups, confirmed: until you ask us to delete them. - Waitlist sign-ups, never confirmed: 30 days. - Waitlist sign-ups, withdrawn: 30 days after you withdraw, apart from a hashed do-not-email record. - Website form service logs: 30 days. - Enquiries: until you ask us to delete them. - Account: until you delete it. ACCESS, CORRECTION AND DELETION ### Ask to see it, fix it or delete it. You can ask what personal information we hold about you and ask us to correct it, by writing to privacy@moulddetect.ai. You can delete your account in the app, or by emailing privacy@moulddetect.ai. We can also delete your backups, waitlist sign-up or enquiries on request. We may need to confirm your identity first. If we refuse a request, we will tell you why and how to complain. COMPLAINTS ### Tell us first; the regulator is there if we fall short. Write to privacy@moulddetect.ai. We will respond within 30 days. If you are not satisfied with our response, you can complain to the Office of the Australian Information Commissioner (OAIC) at oaic.gov.au. CHANGES TO THIS POLICY ### We will tell you before it matters. We will update this policy when our practices change and show the new date at the top. If a change materially affects information you have already given us, we will tell you in the app or by email before it takes effect. CONTACT ### Ask us anything about your data. MouldDetect Technologies Pty Ltd (ABN 80 696 130 784) Privacy: privacy@moulddetect.ai HOW WE HELP Mould Detect gives you tools, information and support for confident decisions. If you suspect a mould issue, use our apps to understand and document what you are seeing, and always seek professional advice. Where possible, confirm with laboratory testing to fully understand your situation. [Join the early access](https://moulddetect.ai/#early-access) --- # Terms of Use | Mould Detect URL: https://moulddetect.ai/terms/ TERMS OF USE ## Fair terms, in plain language. The short version of how Mould Detect may be used while we are in early access. A formal terms document will accompany our public launch. EARLY ACCESS ### We are still building this. Mould Detect products are in early access. Features, availability and the outputs you receive may change without notice while we build, test and learn from real use. That is the trade you are making with us, and we would rather state it plainly than bury it. If something changes in a way that affects you, we will say so. WHAT THE SERVICE IS ### Decision support, not a verdict. Mould Detect is a decision-support and education service. It uses AI-based identification to help you understand and document what you are seeing, so you can decide what to do next and how urgently. It is not a substitute for professional inspection, medical advice or laboratory testing. The full health disclaimer sits on our [privacy page](https://moulddetect.ai/privacy/), in the words it was written in. ACCEPTABLE USE ### Use it on properties you are entitled to assess. Mould Detect is for personal and professional use in assessing properties you own, rent, manage, or have been engaged to inspect. Please do not scrape the service, resell its outputs as your own product, or attempt to degrade, overload or reverse the systems behind it. Sharing a report with your landlord, insurer, property manager or tradesperson is exactly what it is for. YOUR CONTENT ### Your photos stay yours. The photos and property details you submit remain yours, and you grant us only the licence we need to analyse them and return your results. LIABILITY ### Where our responsibility ends. Mould Detect, its parent entities and subsidiaries shall not be held liable for property damage, health deterioration or financial loss resulting from the use or misuse of the data provided by the Mould Detect platform. Users operate this tool at their own discretion and risk. GOVERNING LAW ### New South Wales, Australia. These terms are governed by the laws of New South Wales, Australia. QUESTIONS ### Ask us before you agree to anything. Write to us at hello@moulddetect.ai. A formal terms document will accompany our public launch, and it will replace this page rather than sit beside it. HOW WE HELP Mould Detect gives you tools, information and support for confident decisions. If you suspect a mould issue, use our apps to understand and document what you are seeing, and always seek professional advice. Where possible, confirm with laboratory testing to fully understand your situation. [Join the early access](https://moulddetect.ai/#early-access) --- # Why Mould Detect | The Case for Catching It Early URL: https://moulddetect.ai/why-mould-detect/ WHY MOULD DETECT ## The case for catching it early. Half of Australian homes report mould or damp, and most of it is found late: after the smell, after the cough, after the bond dispute. Mould Detect exists to move that discovery earlier, when it is still cheap and simple to fix. THE NATIONAL PICTURE ### Growing conditions across Australia, right now. Live mould-growth risk for the cities we cover, over the Bureau of Meteorology climate zones that shape it. The same feed powers the homepage outlook. LIVE CONDITIONS UNAVAILABLE RIGHT NOW · THE FEED RETURNS AUTOMATICALLY - LOW - WATCH - HIGH - SEVERE - Climate zones · - Hot humid summer - Warm humid summer - Hot dry summer, mild winter - Hot dry summer, cold winter - Warm summer, cold winter - Mild/warm summer, cold winter VIEW AS TABLE of today's risk by city HOW WE HELP Mould Detect gives you tools, information and support for confident decisions. If you suspect a mould issue, use our apps to understand and document what you are seeing, and always seek professional advice. Where possible, confirm with laboratory testing to fully understand your situation. [Join the early access](https://moulddetect.ai/#early-access) WHAT FEEDS GROWTH ### Mould needs more than moisture. Risk is a stack of factors. The more of them a home has, the faster growth establishes; remove any one and you slow everything down. Humidity in the growth band. Sustained indoor humidity in the high 60s and above is the engine of growth. Holding 30 to 50 percent stalls most moulds. Still air and poor ventilation. Stagnant corners hold moist air against cold surfaces. Cross-flow ventilation and exhaust fans deny mould the still conditions it prefers. Dust as a food source. Mould eats what settles: dust, skin cells, fibres and soap residue. Regular cleaning removes the meal, not just the appearance. Too little sunlight and UV. Direct sun is a free anti-mould treatment: UV damages spores and warmth dries surfaces. Rooms that never see the sun stay colder, damper and darker, which is why shaded walls and built-in wardrobes are classic trouble spots. [Go deeper: the prevention guide.](https://moulddetect.ai/learn/#prevention) THE PROBLEM ### Bigger than a stain. 1 in 2 Australian homes report mould or dampness $28B Annual asthma-related health costs in Australia 24 to 48h How quickly mould can establish on damp material Mould is the new asbestos. MIRKO NAD · CO-FOUNDER THE APPROACH ### Earlier knowledge, better outcomes. A Mould Detect scan gives you a plain-language risk read in about sixty seconds, plus the documentation to act on it. It is not a replacement for professional inspection or laboratory testing; it tells you how urgently you need them, and it favours natural, non-toxic treatment first. Detect sooner. Sixty seconds from photo to read, against days or weeks for lab-only routes. Document properly. Reports formatted for landlords, insurers, property managers and professionals. Resolve naturally. Guidance leads with proven natural solutions and tells you plainly when to call a professional. THE PLATFORM ### Built on a data moat. Behind the app sits a live environmental data platform: weather, humidity, flood and air-quality feeds fused with AI vision trained alongside mycological labs. You can watch it working on the homepage, right now. [See live conditions](https://moulddetect.ai/#live-conditions) ### Know before it grows. Early access is opening now. Scan a wall, keep the record, and find out how urgent it really is. [Join the early access](https://moulddetect.ai/#early-access)