Why the industry that feeds 8 billion people still can’t read its own data

Agricultural data is “fragmented, distributed, heterogeneous, and incompatible.” That’s the verdict from a major Council for Agricultural Science and Technology report published barely a year ago, and it helps explain why AI has struggled to gain traction on farms. Other data-heavy industries, like healthcare or financial services, have established data standards, but agriculture has no universal framework for translating between the dozens of systems that generate field-level information.

This isn’t a new observation, but its persistence is noteworthy. While consumer tech and enterprise software largely solved their interoperability challenges years ago, agriculture still generates enormous volumes of information trapped in incompatible silos. Research institutions publish trial results in inconsistent formats, product manufacturers use proprietary naming systems, farmers record observations with local terminology and retailers track sales without connecting them to agronomic outcomes. The result is an industry sitting on massive amounts of information it can barely use.

“Agriculture doesn’t have a data problem—it has an intelligence problem,” notes Ron Baruchi, CEO of Agmatix, a company building domain-specific AI for the sector. “The data exists. What’s missing is infrastructure that understands what it means.”

According to a McKinsey report, implementing data integration, and connectivity in agriculture could add $500 billion in value to global GDP—a 7 to 9% improvement over current projections. But capturing that value requires solving a problem that general-purpose AI platforms have consistently struggled with.

WHY HORIZONTAL AI KEEPS FAILING IN FARMS

The appeal of applying large language models to agriculture is obvious: A farmer could describe what’s happening in their field and get instant advice on what to do about it, without hiring a consultant or having to wait for a lab. But agriculture’s complexity breaks the approach.

While an LLM trained on internet text might know that nitrogen helps plants grow, it can’t tell you that the right amount changes depending on the growth stage, the soil and what was planted in the same field the previous year. Similarly, computer vision can identify crop stress, but without contextual knowledge of weather, soil and product applications, that insight doesn’t mean much.

You can ask ChatGPT about nitrogen fertilization and get an answer that sounds authoritative. But when you dig into specifics—timing for your soil type, interactions with your previous crop, and product selection based on local availability—the recommendations fall apart.

The same CAST report reinforces this point, noting that many farmers distrust AI because of its “black box” nature—models making predictions without clear explanations behind them. In farming, 90% accuracy on a fungicide recommendation means 10% of the time you’re telling a grower to spray the wrong product at the wrong time.

BUILDING INTELLIGENCE FROM THE GROUND UP

This is where a growing number of companies are taking a different approach—building AI systems designed specifically for agriculture rather than retrofitting general-purpose tools. For example, India-based Cropin, backed by Google, has constructed its own crop knowledge graph spanning 500 crops across 103 countries and recently developed an agriculture-specific micro-language model. Israeli-American startup Agmatix built its own agricultural intelligence system from the ground up—an approach that mirrors, in concept, what Palantir did for defense and intelligence data.

The core of that system is what Agmatix calls “pre-trained ontologies”: Frameworks that encode agricultural relationships before customer data enters the system. Agmatix’s AI engine uses a neuro-symbolic architecture, combining structured knowledge graphs with machine learning. Agricultural relationships—how specific fertilizers interact with specific soils at specific growth stages—are encoded by agronomists, validated through field trials and refined continuously.

What that means, essentially, is that the AI doesn’t start from scratch. Before it touches any farm’s data, agronomists have already taught it how agriculture works—which fertilizers affect which soils, how a crop’s needs change as it grows, and why what was planted last season matters for what’s planted next.

According to the company, the system has structured more than 1.5 billion field trial data points, creating what data scientists call “semantic interoperability”: The ability to translate between different data sources because the system understands what the data means, not just what it says.

But building better technology doesn’t guarantee adoption. McKinsey partner Vasanth Ganesan noted in the firm’s 2024 Global Farmer Insights survey that farmers are “demanding clearer ROI, lower cost of implementation and maintenance and easier-to-setup technologies”—complaints shaped by years of agtech tools that overpromised and underdelivered. A separate McKinsey analysis found that poor user experiences continue to hold back adoption across the sector.

Baruchi says farmers have good reason to be cautious. “Farmers are CEOs operating in one of the most unpredictable industries on earth,” he tells Fast Company. “They balance biological systems, financial risk and environmental volatility every single season. The ROI question is only hard to answer when your platform can’t connect what a grower applies to what actually happens in the field.”

WHERE IT’S WORKING

The approach is already operating across several deployments. BASF has collaborated with Agmatix on digital tools for crop disease detection, including a recently announced project targeting soybean cyst nematode. The company says growers using its prediction platform have reduced fungicide costs by 15 to 20% while maintaining disease control. Its engine is also powering predictive disease-risk modeling in large-scale row-crop systems in the United States.

A national agriculture ministry uses the system to model policy impacts before implementation. On the sustainability front, Agmatix’s RegenIQ platform works with major food and beverage companies to assess which regenerative practices deliver measurable results in specific field conditions—classifying, for instance, Brazil’s 150 coffee-growing localities into six distinct climate clusters, each requiring different approaches.

Cropin, meanwhile, partnered with Walmart in March 2025 to optimize fresh produce sourcing across U.S. and South American markets using AI-driven yield forecasting and crop health monitoring.

THE HARD PART REMAINS

Agmatix represents a broader shift from horizontal AI platforms toward domain-specific solutions. But it isn’t the only company betting that agriculture needs its own AI. John Deere’s acquisition of aerial analytics firm Sentera in May 2025 suggests the industry’s biggest players have reached the same conclusion. The AI in agriculture market is projected to grow from $2.55 billion in 2025 to over $7 billion by 2030, according to Mordor Intelligence. But adoption remains uneven, with 81% of large farms showing willingness to adopt AI, while only 36% of smaller operations plan to do the same.

Agricultural AI adoption is still slow by any standard, and it’s not hard to see why. CAST’s report catalogs the major barriers that agriculture still faces today: High costs, limited rural broadband, insufficient training and unresolved questions about data ownership. These challenges intensify in an industry previously plagued by overhyped technology promises.

But the tailwinds are real. Major food companies have made commitments to decarbonize supply chains that are impossible to fulfill without field-level data. Climate volatility is making predictive tools more valuable. And a decline in U.S. public agricultural R&D spending — down roughly a third from its 2002 peak, according to USDA data — is creating a vacuum that private-sector platforms are positioned to fill.

The question isn’t whether agriculture needs better data infrastructure. It’s whether the companies building it can survive farming’s patient adoption timelines long enough to reach critical mass and whether the benefits will extend beyond the largest farms that can already afford to invest. For an industry responsible for feeding 8 billion people, getting that balance right matters enormously.

View the full article