SukhaRakshak AI — Drought Intelligence
Agriculture & AgriTech
AI that makes sense of drought at scale: reasoning across drought indices, satellite data, forecasts and contingency plans to give decision-makers current, evidence-backed answers as droughts spread.
SukhaRakshak AI, built by IWMI, is India's first AI-based system to tackle drought. Drought now touches many regions at once, and every decision depends on combining drought indices, satellite data, forecasts and district contingency plans correctly and in time. Kenpath is IWMI's technology partner for the next stage, using AI to reason across that evidence so the platform can serve many more places and decision-makers.
How AI helps us tackle the complexity
Evidence that is always current
Live connections to every data source mean decisions reflect today's conditions, with a clear "as of" for each source.
One picture from many sources
Drought indices, satellite observations, forecasts and contingency plans are combined into a single, consistent view for any block, district or state.
Questions reasoned through, not looked up
Multi-step questions, like comparing seasons or finding the worst-hit blocks, are planned and worked through, and a short question is asked when something essential is missing.
Every recommendation traceable
Each statement links to the evidence behind it, and the system declines rather than guesses when evidence is thin, so officials can stand behind what they act on.
Built to scale across states
Open standards, a model-agnostic core and a sovereign hosting option let new states join by configuration, not rebuilds.
Measured, then handed over
Every improvement is checked against a baseline agreed at the start, and all code, models, prompts and documentation are handed over to IWMI.
Where the complexity sits, and where AI helps
The pilot showed that real users get real value from SukhaRakshak AI and that its data pipelines run. The upgrade puts AI where drought is hardest to handle: reading and combining evidence that arrives from different sources, on different schedules, at different scales; reasoning through questions that need several steps; and showing the evidence behind every recommendation so it can stand up in official decisions. The answer adapts to the person who needs it, plain and actionable for a farmer, with depth and evidence for an extension officer or a district manager. Because droughts now reach more states, the platform is built to be adopted by configuration and measured at every step against a baseline agreed at the start.
Kenpath's role
Kenpath is the technology partner for the upgrade. We work alongside IWMI's SukhaRakshak AI team on IWMI's own infrastructure, so that IWMI and its partners can run, extend and take the platform to new states on their own. Code, models, prompts and documentation are delivered to IWMI, with knowledge transfer built into the work.