Search the GenAI Directory

Measuring wheat plants by cutting them up destroys the plant and the trial. This is code plus a multi angle image dataset for rebuilding wheat plants in three dimensions from photos, so you can measure them without touching them.

Your organisation knows it should be doing something with AI but cannot say where it currently stands. This is a published ladder of stages with survey findings and self-placement questions, so a team can locate itself and pick a next step.

Irrigation planning and water accounting need to know how much water crops actually used, not just how much rain fell. This is a monthly satellite-derived record of water use by Australian farmland, ready to download.

Tuning a model for Brazilian farm questions needs example exchanges you probably do not have. This is a downloadable dataset of agriculture question and answer pairs drawn from government agricultural material.

Coffee growers have questions between visits from a technical adviser. These are two Spanish language chat assistants for coffee production, documented in a Colombian coffee research institute's technical note series.

A small natural food brand cannot afford a full analytics team but still needs to know how it is selling. This is a hosted service that answers category, retail and promotion questions from retail data and sends a monthly interactive report.

Comparing how much livestock African countries produce means digging through official reports. This is a downloadable table of livestock production figures by country, ready to load into a script or notebook.

An agronomist covering many fields cannot walk every acre or remember every scouting note. This is a hosted assistant for retail agronomy teams that draws on close-up crop imagery and written notes to raise field issues and follow-up opportunities.

Planning a crop season or trying to work out what is eating your plants often means waiting on an adviser. This is a chat-style farming assistant in a web browser that talks through crop plans, pest questions and regenerative practices.

You need draft farming text for an app or experiment and a general model keeps missing the vocabulary. This is a downloadable set of model weights, a small older text model further trained on agricultural writing, that you wire into your own code.

Swedish farmers asking questions in their own language get thin answers from general assistants. This is a hosted assistant built on a customised language model, mixing selected farming references with documents the farmer uploads about their own operation.

Farm back office work such as environmental paperwork, input orders and co-op reporting swallows staff time. This combines an agent platform with a delivery team, wiring agents into existing farm business systems and public data services.

Shoppers increasingly ask AI assistants about products, and you cannot tell whether your brand comes up. This is a hosted platform that tracks how brands appear in those assistants and reshapes product data into machine readable web content.

Your team wants a farmer facing advisory assistant but has nobody in house who has built one. This is consulting and build work, where specialists scope the advisory idea and deliver working automated workflows with you.

Comparing agricultural models is awkward when everyone writes their own scoring scripts. This is an open code repository holding evaluation material and scripts for assessing farming focused models.

Field operation files from one brand of equipment rarely open cleanly in another company's software. This is an open code framework and shared data model for translating farm machinery and field records between systems.

Plant image work often stalls because you cannot find annotated pictures covering weeds, cover crops and cash crops together. This is an open image repository with labelled plant photographs from both controlled plots and real fields, reachable through code.

Agronomy trial and crop nutrition results are often spread across files, so answering a practical question takes manual digging. Agmatix Insights, previously surfaced as Leafy, is a hosted GenAI workspace that analyzes harmonized field-trial data and turns it into cited agronomic findings and product-performance summaries.

Comparing results from several crop models is painful when each one wants its own file layout. This is a collection of converter programs that translate model inputs and outputs between formats so runs can sit side by side.

Training a farm vision model means hunting down datasets and writing loaders for each one. This is an open Python library that lists, downloads and prepares agricultural image datasets, runs comparisons and can make synthetic images with labels.

Before trusting an assistant with farm questions that involve photos, you want a fair way to test it. This is a test set of agricultural questions combining text and images, with expected answers for scoring models.

Agronomic advisors and growers often spend hours cross-referencing imagery, soils, weather, input applications and yield history before they can answer one field question. The AGMRI AI Agent is a live conversational assistant inside the commercial AGMRI platform that analyzes connected field data and produces field-level answers, reports and prescription outputs.

Driving straight lines and switching sprayer sections by eye wastes seed, fuel and chemical. This is open source software for field mapping, section control and tractor guidance that you install and run yourself.

Retail agronomists switch between grower records, product data and field information to answer a single question. This is an assistant for agricultural retailers that pulls together farm data services and the dealer's own records in one place.

It is hard to tell whether a model really reasons about farming or just repeats textbook lines. This is a small test set of open-ended agricultural questions written by specialists, used to compare how models answer.

Farm questions often mix soil, watering and European subsidy paperwork, and the answers live in different places. This is a preconfigured chat assistant reached by link that discusses farm management, soil, irrigation and European farm policy rules.

Deciding when to switch irrigation on takes constant attention to soil, weather and crop stage. This is a hosted system of cooperating software agents that reads those conditions, works out watering needs and can trigger the irrigation itself.

Farmers and advisors ask the same crop questions over and over, and expert time is short. This is a language model already adjusted on farming material, offered as downloadable weights with code so you can build an assistant.

When you tweak a model for farming questions you need a quick way to see whether anything broke. This is a small downloadable test split of agriculture questions you can run a language model against.

Farm questions come up at odd hours when there is no adviser to call. This is a chat assistant on the web that takes general farming and agronomy questions and answers them in conversation.

Trying several satellite models for field boundaries usually means learning several different interfaces. This is an installable toolkit that puts a number of open satellite models and image representations behind one common set of calls.

An AI assistant asked about spray timing or water use has no way to work out the farming numbers itself. This is a code library and plug-in interface for AI assistants that supplies weather, water use, heat accumulation, spray windows and soil figures.

General chat assistants know nothing about your particular paddocks. This is a hosted assistant that answers agronomic questions against your own field boundaries, satellite readings, forecasts and disease risk, and takes spoken notes on field work.

Advisors would like to show a crop photo and simply ask what is wrong with it. These are downloadable weights adapting an open model that reads text and images, shipped with a simple chat interface and scoring scripts.

Image models trained on everyday photos often miss what matters in crop and animal pictures. These are downloadable model weights and pretraining data that link farm and livestock images with text descriptions, for you to build vision tools on.

Judging whether a model can reason about pest and disease pictures in Chinese needs a proper test set. This is a Chinese language test collection pairing thousands of crop images with questions about pests and diseases.

Questions about crops, livestock and running the place come up between visits from anyone who can answer them. This is a preconfigured chat assistant, opened through a shared link, covering crop production, farm management and livestock topics.

You want quick thoughts on soil condition and getting more from a field without booking a consultant. This is a preconfigured chat assistant, shared as a link, aimed at growers and crop advisers asking about crops and soil.

Watching crop prices and coordinating several small farm bots usually means stitching scripts together. This is a plug-in interface for AI assistants that supplies current crop prices and helps run farming focused bots.

Anyone watching harvests and supply worries has to pull together prices, weather and crop condition reports. This is a hosted plug-in interface that gives AI assistants commodity, climate, crop risk and food supply information in one place.

You want a live software copy of a farm holding to try changes before making them for real. This is an open platform for building such virtual copies, which you can extend with your own simulation and analysis tools.

Margins, input costs and market conditions raise questions that need an economics angle rather than an agronomy one. This is a preconfigured chat assistant, opened by link, for discussing the economics of farming and farm trade.

Working out which crop is growing where from imagery needs example data to learn from. This is a downloadable dataset for tracking fields and sorting crop types, also reachable through a data site interface.

Search and text tools often miss the specialist vocabulary used in farm research papers. This is a set of numerical word representations trained on agricultural research writing, offered through a website for researchers and developers to build on.

Your app needs to know what is growing in a field and what the surrounding land looks like. This is a web connection point that returns structured field level crop and landscape information for use in software or models.

Guessing what a field will produce shapes decisions on selling, storage and inputs. This is a listing in a cloud provider's agent directory for a hosted tool that looks at yield, aimed at growers and their advisers.

Guessing how much a field will produce is easier with past records to learn from. This is a downloadable dataset of crop yield figures you can use for training, grounding or checking a model.

Farming and agribusiness questions come up and you want a knowledgeable place to start. This is a preconfigured chat assistant, opened through a shared link, discussing farming practice and the business side of agriculture.

Paperwork and repeated handoffs across farm and supply chain jobs still soak up staff hours. This is a build and integrate service that sets up assistants and automations on common chat and workflow tools for those tasks.

You need a chat assistant that can answer farmer questions and collect field details, without building one from nothing. This is a configurable template for that kind of assistant, ready for your own crops, wording and data fields.

You have an idea for an agricultural software tool but no in house AI engineers to build it. This is a development service where an outside team designs and builds custom AI applications for farming clients.

Training farm specific models stalls when nobody has labelled the images, text or records you hold. This is a service where people label your agricultural data and check each other's work before it goes to training.

Early experiments often need a starter set of farm data rather than a bespoke collection. This is a farming dataset published in an open catalogue, downloadable for training a model, giving it reference material or scoring results.

Sometimes you just need general farming records to try an idea before investing in data collection. This is a publicly posted farming dataset that can serve as training material, test material or background reference.

You have farm data but no engineering team to turn it into something an assistant can use. This is a data and AI engineering service for agriculture, covering pipelines through to assistants that take actions.

Extension teams answer the same farming questions over and over, and writing them up takes time. These are downloadable model weights, with code, for a text model tuned to write and answer common agricultural questions.

Getting useful answers from a chat assistant about farm planning depends heavily on how you ask. This is a library of reusable prompt templates for agricultural planning, analysis and advisory work, offered as a listing or with setup help.

Sometimes you just need a small model that produces farming flavoured text for a draft or an experiment. These are downloadable model weights, with code, for an older text generator further trained on agricultural writing.

Finding out what farming models already exist is a research project in itself. This is a curated public list of agricultural language, image and combined models along with applications built on them, kept in a code repository.

Rolling out AI in agronomy raises awkward questions about how you will check it and who signs off. This is a consulting workstream that designs farm specific tests and oversight rules around an AI deployment.

Staff waste time hunting through internal agronomy documents for answers that already exist somewhere. This is a build service that creates an internal search assistant over your own farming knowledge base.

New growers often have plain practical questions and nobody nearby to ask them of. This is a preconfigured chat assistant, opened through a shared link, that talks through everyday farming decisions in a mentoring tone.

Farmers' questions pile up faster than advisors can reply to them. These are downloadable model weights, with code, for a language model adjusted to answer agricultural questions inside your own application.

When you want a working farm question and answer bot to show colleagues quickly, starting from a blank page slows everything down. This is a deployable code template that searches an organic farming knowledge base before replying, using Amazon's hosted models.

Building or checking a farm advice assistant goes faster with ready made examples. This is a downloadable set of agricultural questions with matching answers, usable for tuning a model or scoring one.

Tuning an assistant to answer grower questions needs example questions with answers attached. This is a downloadable set of English farming question and answer pairs you can train on or use for checking results.

Turkish speaking growers ask questions in their own language, and most training material is not. This is a downloadable set of farming questions and answers translated into Turkish, meant for building assistants that handle more than one language.

Teaching or learning farming topics goes better with something that adjusts to the learner's level. This is a preconfigured chat assistant, shared by link, covering crop care, pests, livestock and newer farm technology as lessons.

Farm advice written in Hindi needs translating into other Indian languages, and general translation data rarely covers farming terms. This is a downloadable collection of matched sentence pairs on crop practices, weather advice and government schemes for training translation models.

Advice locked inside training videos is hard to search or reuse. This is a downloadable set of question and answer records pulled out of farm advisory videos, ready for training or grounding a model.

Video is a poor format when you need to search what was actually said. This is a downloadable collection of written transcripts from farm training videos, useful as background text for a model.

Before trusting an AI system on a farming or food task you need a way to check it. Semantic segmentation of six aerial farmland anomalies including weed cluster, standing water and planter skip, scored by mean intersection over union.

Trying out an analysis method on real trial results means first finding real trial results. This is a data package for the R language holding many agricultural datasets, from small plot and multi site trials to uniformity trials and yield monitor records.

Digging through agricultural literature and sources for a research question can take many hours. This is an installable add-on for the Claude assistant that breaks longer agricultural research work into ordered steps.

A sick looking crop needs naming before the right treatment can be picked, and a plant doctor is not always nearby. This is an assistant that takes photos and questions, sorts them to the right checker, and pulls answers from farm reference material.

Reading through service documentation and guessing at the plumbing slows down a first working prototype. This is a set of ready-made example pipelines showing how to call an agricultural image service for plant and picture comparison tasks.

Chinese language farming questions are largely missing from mainstream model tests. This is a Chinese agriculture test collection covering a range of farming tasks, downloadable for scoring language models.

Pretraining a vision model for agriculture means gathering field pictures from many scattered public sources. This is a downloadable image collection assembled from public field photography, prepared for further training of picture models.

Tracking crops across a season means stitching together pictures from several satellites at different scales. This is a downloadable model trained on satellite image sequences over time, meant to be adapted for your own agricultural mapping work.

Before trusting an AI system on a farming or food task you need a way to check it. Catalogue of 311 European services to validate, test and evaluate agrifood products, including AI model evaluation, explainability audits and safety qualification.

Many farmers already use messaging apps all day and would rather ask a question there than open new software. This is a chat service on WhatsApp that replies to farm questions with answers shaped around your region and the crops you grow.

Farming questions need a model that has read agricultural material, and there is little to test one against. This is a research release bundling a farm focused language model with its training dataset and a set of test questions.

Traders need a view on where prices might go and what to do about it, pulled from scattered sources. This is a hosted market intelligence platform where several language model agents produce forecasts, trade ideas and written scenario reports.

Farmers who prefer another language often cannot get advice in it at all. This is a preconfigured chat assistant, opened through a shared link, answering farm questions in several languages with an emphasis on sustainable practice.

You want a farming question and answer assistant you can inspect and change yourself. This is open code that pairs a freely available language model with a farming reference collection, searching that material before it answers.

Getting a model to link farm pictures with the right words takes matched training examples. This is a downloadable dataset about crop pests and diseases, published for people training or testing their own models.

You want a model to follow instructions about pests and plant health rather than answer vaguely. This is a downloadable dataset of instruction style records on crop pests and diseases, with code alongside it.

Teams building assistants that answer questions about farm photos need a shared yardstick. This is a published test set of agricultural images with questions and reference answers, for scoring models that handle text and pictures.

You have built a farm chatbot that handles pictures and now need a way to judge its replies. This is a published set of questions and reference answers covering text and images for farming.

Teaching a model to talk through a crop problem from a photo needs examples of that conversation. This is a downloadable collection of leaf and crop pictures paired with back and forth diagnostic dialogue.

You want to know which language models actually hold up on farming questions before you build on one. This is a hosted scoreboard page that shows how different models score on a shared set of agriculture test questions.

Before you rely on a farm assistant, you want a fixed set of questions to score it on. This is a downloadable evaluation set of 800 agricultural question and answer records for measuring model replies.

Writing farming content or long answers with a general model can feel off in tone and detail. These are downloadable model weights, with code, for a larger language model adapted to agricultural material.

Getting enough farming text together to train or ground a model is a chore on its own. This is a downloadable body of raw agricultural writing, bundled so you can feed it straight into training or search.

Adjusting a language model for farming needs a lot of worked examples, not a handful. This is a downloadable training set of roughly 146,000 agricultural instruction records, with code to help you use it.

Testing whether a model really understands crop photographs needs more than a handful of examples. This is a downloadable collection of farm images paired with questions and answers, gathered from many existing datasets, plus scoring scripts.

Building a farm advice bot for Indian users means starting from a model that knows local crops and local questions. These are downloadable model weights, with code, for a small language model tuned on Indian agricultural questions and conversation.

Before trusting an AI system on a farming or food task you need a way to check it. Predicts the five-class ordinal viticulture potential of each land pixel from multi-temporal Sentinel-2 imagery over southern France.

Checking whether an assistant gives sensible farming answers needs a big pool of questions to test against. This is a large collection of agricultural question and answer pairs covering crops, fisheries, livestock and farm inputs.

Before you trust an image model on real field conditions, you need messy pictures to test it against. This is a downloadable set of drone photographs of crops, put together as a test collection for vision and image search models.