Search the GenAI Directory

Plant floor teams have production data in one system and their work records in another, so questions take digging. This is a hosted assistant over those factory and business systems that answers questions and sets up follow-up work.

A grower with an arable crop question can lose an afternoon digging through decades of research booklets. This is an assistant on a research organisation website that searches its own arable crop material before answering and links to the source documents.

Food and drink companies need an answer on a labelling or regulatory question without waiting on a consultant. This is a subscription assistant, trained on regulatory sources, that answers compliance questions and links to the rule it is drawing on. It is sold in two plans, a standard one with a monthly chat allowance and a Pro one without, and it also appears inside other Trustwell products.

Questions about land and maps often need someone who knows mapping software. This is a hosted assistant, listed for the farming industry in a cloud marketplace with a trial, that answers questions about location and imagery data.

Before trusting an AI research helper with your science work, you want to see how it does on set tasks. This is an open code project that tests research assistants against scientific tasks and publishes scoreboards of the results.

Grain trading buries staff in certificates, contracts and rule checks that all need reading by hand. This is a hosted platform of chat driven assistants that scan documents, watch compliance and can also be wired into other AI tools.

Farm accounts and production figures get reported every month, but somebody still has to write up what the numbers mean. This is a feature inside a farm financial reporting tool that drafts that written commentary and updates it as new actual figures arrive.

If you want to repeat someone else’s model comparison rather than take their word for it, you need the run itself. This is a published evaluation run comparing language models on agronomy questions, with the data to reproduce it.

A released, open ROS 2 framework for building physical AI agents that combine LLM, VLM, speech, memory and robot components. The publisher documents fully offline local models and edge optimization for NVIDIA Jetson.

Daily calls on labour, deliveries, production and stock levels eat up management time and often get made late. This is a hosted set of trained software agents and an optimiser that carry out those operational decisions for food and farm businesses.

Answers your team needs are buried in manuals, trial reports and spreadsheets that no chat tool can reach. This is a connection point and knowledge base setup that searches your documents before answering, offered with a farming use case.

When you want to prototype a farm advice chatbot, starting from an empty folder is slow. This is an open code repository for a farming focused generative model, with an interface you can call from your own app.

Your team wants to build a farm data assistant but nobody wants to design the architecture from a blank page. This is published reference material and deployable templates showing one way to assemble such an agent on a cloud platform.

You want to see how a shopping and cooking helper that reads product labels could be put together. This is an open sample code repository for a food and nutrition app, built on cloud services and language models for developers to copy.

Wiring an assistant into a farm data platform means learning its interfaces before you write a line. This is an installable skill file that teaches a coding assistant how to work with Microsoft’s agriculture data service.

Advice about animals goes wrong quickly if a model’s basic animal science is shaky. This is a multiple-choice test set on animal knowledge, including farmed species and animal science topics, for checking language models.

Grain originators juggle calls, contracts and follow-ups with a long list of farmers. This is a helper inside a grain merchandising platform that keeps track of those producer relationships and moves the routine steps along.

Grain price moves matter to your model or dashboard, but the public reports are awkward to pull in. This is a commercial data interface that serves grain prices published by the US Department of Agriculture along with related market figures.

Livestock records, treatments and notes tend to live on paper or in your head until something goes wrong. This is a farm and ranch management app that keeps animal records, watches for health signals and answers questions about your own data in plain English.

Checking what your irrigation controllers are doing means logging into their own dashboard. This is a read only plug-in interface for AI assistants that exposes a dozen tools for reading data from one brand of irrigation controller.

Testing early product ideas with real shoppers costs money and weeks you may not have at concept stage. This is a hosted service that builds simulated shoppers from household purchase records, then returns ratings, written comments and suggested tweaks.

Biological soil test results usually arrive as reports that other systems cannot read. This is a service other software can call to pull soil test records, reports and biological activity results into your own platform.

Sheep and beef farmers often know the guidance exists somewhere but not which booklet it is in. This is an assistant built on a farmer organisation’s knowledge library that answers questions and drafts plans and checklists for your situation.

Lab records for seed and trait work end up scattered across notebooks, so simple questions take hours of digging. This is an assistant inside a lab record system, plus a plug-in interface that lets approved AI assistants read experiments, samples and results. The publisher describes it in the context of Benchling Assistant + agents + Benchling MCP server.

Pulling world food and farming statistics out of the United Nations database usually means wrestling with query parameters. This is a plug-in interface for AI assistants that lets you ask for those statistics in ordinary language from tools like Claude or Cursor.

You want a farm question answering feature without training a whole model from scratch. These are downloadable adapter weights, with code, that adjust an existing text model so it answers agricultural questions.

Brazilian producers want somewhere to ask a quick farm or money question at no cost. This is a free assistant on WhatsApp answering those questions in Portuguese, from the same group behind a farm credit analysis agent.

Studying maize plant shape usually means destructive sampling or slow manual measurement. This is a downloadable collection of three dimensional point scans of field grown maize plants from a diverse set of lines.

Farm advice services in India work in many languages, and testing usually only covers English. This is an agriculture test set written in Indian languages, part of a wider family of language benchmarks.

Training image models needs labelled pictures, and collecting enough of them in the field is slow. This is a code pipeline that builds rendered scenes with labels already attached, using open source three dimensional graphics software.

A released ROS 2 node and agent runtime for connecting OpenAI-compatible LLM or VLM backends to robot topics, tool calls, skills and local hardware. It can use local or hosted model servers and publishes multimodal and reasoning streams inside ROS.

Before trusting an AI system on a farming or food task you need a way to check it. Leaf instance segmentation on photogrammetric point clouds of real sugar beet breeding trials covering 48 plant varieties.

You want a plant DNA model sized to the hardware you actually have. These are downloadable plant genome models in three sizes, trained on forty three plant genomes, meant as a starting point for your own tasks.

Writing up the same category and brand analysis for account after account eats an analyst’s week. This is a hosted feature inside a retail measurement platform that reads your own and third party sales data and drafts the write ups. The publisher describes it in the context of Circana Liquid AI.

Breeding programmes keep seed stock, trial and measurement data behind a standard interface that assistants cannot read on their own. This is a plug-in interface for AI assistants that opens up that data so they can work across varieties, trials and observations.

Breeding programmes each store trial, genetic and seed stock data differently, so tools do not talk to each other. This is a published interface standard with a test server, so plant breeding software can exchange those records in an agreed shape.

Preparing for a retailer food safety audit means interpreting a standard written in dense language. This is a preconfigured chat assistant, shared as a link, that answers questions about achieving and maintaining BRC Global Standards.

Breeding databases each store trait and genetic data their own way, so moving records between them is a chore. This is an agreed set of web connection points that breeding systems use to exchange plant trait and genetic data.

Wine and hospitality businesses hold years of material in documents nobody outside the office can query. This is a private assistant trained on your own content, branded as yours, with several languages and drafted social posts included.

Brazilian farm figures are spread across many public sources, each with its own quirks. This is a plug-in interface for AI assistants, with code you can inspect, that reaches ten public Brazilian agricultural data sources.

Farm records kept in a farm management app are hard to reach from an assistant. This is a plug-in interface, currently private, built to connect AI assistants to an open farm management system.

Wiring your own software to a field robot usually means working out its controls by trial and error. This is a documented command interface, over a cable or a network connection, for instructing Burro robots and reading their state.

Testing whether an assistant answers German food safety questions properly needs a shared set of questions. This is a downloadable German language question and answer collection built from a federal food safety agency’s published documents.

Preparing a retailer or category review deck eats a morning that a small brand team does not have. This is a feature in a retail data platform that assembles a shareable slide deck from your sales numbers.

A gated, multilingual CABI corpus on palm weevils, including management and bioprotection context, structured for LLM and retrieval-augmented agricultural advisory applications.

A gated, multilingual corpus of CABI plant-health advisory content structured for use in large-language-model and retrieval-augmented-generation applications.

Your organisation’s research reports go unread because finding the right passage inside them is too much work. This is a branded assistant platform that searches your own documents before answering and cites them, deployed inside your cloud account. The publisher describes it in the context of embeddable AI knowledge platform.

You want environmental and crop modelling inside your own product without building the models yourself. This is a modelling service delivered either as a tailored platform or as an interface your software connects to.

Grain traders lose time clicking between price screens, news feeds and chart tools to answer one question. This is a chat helper inside a commodity trading platform that answers market questions in plain words and sets up the screens you asked about. The publisher describes it in the context of AI assistant in cmdtyView.

Before trusting an AI system on a farming or food task you need a way to check it. Identifies the type of disease present on a cassava leaf image.

Counting or tracking cattle in photos and video needs annotated examples to get started. This is a publicly shared image dataset with cows marked out, suitable for training or checking detection models.

Seeing how cattle numbers in African countries have shifted over the years means stitching sources together. This is a downloadable table of cattle head counts by country and year, ready to load into code.

You need a long run of cattle numbers for Asian countries without hand copying from reports. This is a downloadable year by year table of cattle head counts, formatted so a script or model can read it.

Comparing herd sizes across European countries over time is fiddly when each source differs. This is a downloadable table of cattle head counts by country and year, tidied for direct use in analysis.

Watching dairy cows all day to log feeding and chewing is not practical for a research team. This is a downloadable video collection labelled for five cow behaviours, set up so you can compare models on it.

Research on how a changing climate hits farming and food supply needs weather layers past and future. This is a downloadable set of historical and projected climate data prepared for agricultural and food security work.

A farmer asking a question out loud in Kikuyu or Hindi cannot reach an English only AI service. These are downloadable speech recognition and translation models, plus their training data, tuned on farming questions so local language queries can be passed on.

Turning recorded farm training into something a model can learn from takes work. This is a downloadable set of question and answer records drawn from agricultural videos, for training or for checking answers.

Spoken advice in farm videos cannot be searched or quoted easily. This is a downloadable set of written transcripts from those recordings, republished so a language model can draw on the text.

Growers and advisers often just want a straight answer to a crop or soil question. This is a small hosted demonstration app, with the model files also published, that takes agronomy questions and replies to them.

Before trusting an AI system on a farming or food task you need a way to check it. Assigns probabilities across five smallholder crop damage categories from field photographs, scored by log loss.

Farm research findings sit locked inside thousands of separate documents. This is a downloadable collection of agricultural research texts cleaned up so a model can read them or search them before answering.

Farmers with a question often have nobody to ask, and extension staff are stretched thin. This is a simple hosted demo of a farming question and answer bot, with the underlying model also available to download and run.

Sometimes the answer you need is buried in a set of food and nutrition documents rather than in your head. This is a small hosted web demo, with its files published too, that searches a document collection and then answers your question.

Answering a question from the farm research literature means opening report after report. This is a downloadable body of agricultural research documents prepared as plain text for model training or document search.

Food policy research is spread across long reports that are slow to comb through. This is a downloadable English language slice of an agricultural research document collection, packaged for model training and document search.

Food policy research sits in long discussion papers that few people have time to read end to end. This is a small hosted web demo that searches a collection of those papers and answers questions from what it finds.

Spotting something wrong with a plant and not knowing what it is means guessing or asking around. This is a small hosted web demo that looks through plant health reference material and answers questions about pests and diseases.

Checking an assistant that searches research documents before answering needs questions with known answers. This is a question and answer set drawn from agricultural research literature, arranged so questions and source passages do not overlap.

Testing a document-searching assistant also means checking cases where the answer sits plainly in the retrieved text. This is a companion question and answer set from agricultural research writing, where questions and source passages overlap.

Agricultural library holdings are useful only if a machine can actually read them. This is a downloadable set of English agricultural research documents converted into text you can train on or search.

Australian dryland growers and their advisers want answers grounded in local trial work, not overseas guidance. This is a chat assistant built on decades of regional farming systems research that covers topics like sowing timing, nitrogen and gross margins.

Crop protection chemists need new molecule ideas and a way to filter them before lab work starts. This is a hosted molecule design platform with generation, scaffold swapping, property tuning and a chat assistant, now used in one company’s pesticide discovery work.

Growers who work in Chinese need answers in Chinese, and most question sets are in English. This is a downloadable set of Chinese language questions and answers covering farming, forestry, animal keeping and fishing.

Orders reach food distributors as voicemails, photos, texts and scribbled notes that staff retype into the order system. This is a hosted agent that turns those messy messages into tidy order records and can post the checked ones by itself.

When a grower or advisor asks a farming question, sometimes a general assistant is what you reach for. This is a general purpose language model, available through a programming interface, that has been tested on farm advisory question sets.

You may want to see how a large general assistant handles crop and farm management questions before building anything specialised. This is a general purpose language model, reachable through a programming interface, that has been scored on an agricultural advice test set.

Planning around weather risk needs past records plus outlooks for the weeks and seasons ahead, in one place. This is a data service other software can call for historical, short range, seasonal and long term climate figures with crop data.

You need a grower’s field data in your own app, with their permission and without exports. This is a developer interface, protected by a sign in and approval step, for reading farm and field records held in a crop platform.

You want to try practices that hold up in a changing climate, but the advice is scattered around. This is a preconfigured chat assistant, opened by link, that works through sustainable farming practices in an ordered way.

Commodity trade admin, from contracts to shipment paperwork, swallows time between deals. This is an assistant layer added to a commodity trading and management system that takes on parts of those routine workflows.

Questions about compost and soil amendments come up constantly in advisory work, and general models often have thin coverage. This is a part-generated body of composting text prepared as reference or training material for farm assistants.

Knowing whether cattle are grazing, lying down or standing normally means someone watching the paddock. This is a downloadable set of pasture photos with boxes drawn around animals and their behaviour labelled.

Herd and milking data from a dairy platform is hard to feed into your own analysis or models. This is a web connection point that pushes and pulls dairy records so they can flow into your data pipelines.

If you are building a farming chat assistant, starting from nothing means rebuilding plumbing others have already written. This is an open code repository for an agricultural chat project that you can read, run and adapt in your own work.

Weather and climate layers for a crop model are scattered across archives that are slow to work through. This is a Python interface for downloading past and forecast climate data from a European public data store, straight into your code.

Starting a farm assistant from a blank page means solving the same plumbing problems everyone else already solved. This is a set of ready made assistant templates you deploy in containers, wired to query farm data and draft planning notes.

Growers experimenting with natural farming methods have few places to ask practical questions. This is a small hosted chat demo, with its files openly published, that talks through natural farming practices and inputs.

Machine and field records from a manufacturer’s operations centre are useful only once something can read them. This is a small plug-in interface that connects AI assistants to a tractor maker’s farm operations data.

Producing video that follows a given scene layout is hard without a model built for it. These are downloadable weights for a video generating model that takes a written description plus guide videos such as depth or outlines.

Robot vision teams never have enough footage of unusual field situations. This is an open recipe with code for further training a video generating model on captured crop rows, so you can produce new synthetic scenes.

Sorting out which weed is which in a cotton field is a job growers and researchers repeat constantly. This is an image dataset covering fifteen weed types found in cotton, bundled with code for training and evaluation.

Telling one cow from another in camera footage is hard without example images per animal. This is a large downloadable image collection of individual cattle, gathered for work on recognising animals by sight.

Understanding how roots and shoots grow together, and how roots meet the soil, is hard to observe directly. This is simulation code that models plant structure above and below ground along with soil interaction.

Store level sales data changes daily and nobody has time to read all of it, let alone act on it. This is a hosted group of assistants that watch retail sales figures, flag openings, and draft charts, slides, emails and follow up tasks.

Managing seasonal farm labour paperwork means logging into yet another system to check details. This is a plug-in interface that lets chat assistants such as ChatGPT and Claude read and act on a farm labour management platform.

Telling crops apart from weeds in field photos, species by species, needs labelled pictures at scale. This is a large downloadable image dataset with seventy four crop and weed classes marked, for training and checking field vision models.

Telling a crop seedling from a weed in a photo is the first step before spraying decisions. This is an openly listed asset for crop and weed detection, offered through a website for training or scoring models.

Crop scientists want to know whether a language model is actually useful for their work before relying on it. This is a test and training bundle of crop science tasks with material for scoring model answers.

A public MCP server and retrieval layer over US row-crop pesticide, herbicide and fertilizer labels. It provides label-rate, re-entry interval, pre-harvest interval, rotation and tank-mix retrieval to the Drawbar advisor.