Why KrishiAI wants to build one AI system for the entire farm

Nagpur: For years, agri-tech has often grown in pieces. One application diagnoses a crop disease, another provides weather information, another helps with markets, while schemes, soil and livestock can sit in entirely different digital silos. For the farmer, the problem is obvious: the farm does not operate in silos.

That fragmentation is the starting point for KrishiAI’s proposition. The Nagpur-based startup believes the next generation of agricultural technology should not ask farmers to find a different app every time a new problem appears. Instead, it is building Vasundhara AgriOS as a common operating layer for the farm, bringing multiple forms of agricultural intelligence into one system.

Abhilash Khopde, Director at KrishiAI, says the thinking behind the initiative was to deliberately move away from the prevailing model of building separate digital products for separate agricultural problems. “The company’s approach is to make the different capabilities work together, so that an answer about irrigation, for instance, can take into account the farm’s soil, crop stage and weather rather than treating the question in isolation,” he says.

Shweta Kirad, Director at KrishiAI, points to the farmer’s experience at the centre of this approach. The aim, she says, is not to make technology more complicated, but to “make it simpler for the farmer by bringing the intelligence to one place and allowing the interaction to happen in familiar languages.”

Vasundhara AgriOS currently uses 28 registered AI agents covering 27 capabilities, including crop planning, disease diagnosis, soil interpretation, irrigation, livestock, market and money, and government-scheme eligibility. A digital farm record called Krishi Pratiroop is designed to retain information on soil, crop stage, water, weather and farm history, allowing the system to build context across a season.

The platform is already being tested in Maharashtra’s Nagpur and Bhandara. As of 19 August 2026, KrishiAI reported 202 farmers, 128 mapped land parcels and about 554 acres mapped through GPS, alongside 4,646 scheme-eligibility assessments across 23 central and state schemes. Marathi, Hindi and English are currently live, with the system designed for entry-level Android phones and intermittent rural connectivity.

The company is also clear that this is still an early-stage experiment. Feature-phone access through IVR remains on the roadmap, while its proposed developer platform and IoT hardware programme are at different stages of development. The startup is being mentored as consultants by Sumit Pathak and Satwik Deshpande.

For KrishiAI, the larger bet is that agriculture needs an ecosystem rather than another standalone app. If the approach works, the farmer should not have to think about which technology solves which problem. The system should understand the farm first, and then bring the relevant intelligence to the farmer.