Kisan Sahayak

Kisan Sahayak is a voice-native AI assistant built to make agricultural advisory services accessible to Indian farmers who may not read, type, or speak a single fixed language. Farmers call in and describe their problem naturally — in Hindi, English, or a mix of both — the same way they would talk to a person, not the way they'd fill out a form.

Instead of jumping straight to an answer, the assistant asks only the follow-up questions it actually needs: crop type, symptom description, how long it's been happening, and how much of the field is affected. It remembers everything said earlier in the session, so farmers never have to repeat themselves, even if they switch topics and come back later.

Once enough detail is gathered, the assistant checks a verified agricultural advisory knowledge base and grounds its response in real data rather than generating advice on its own. If it finds a confident match, it explains the likely cause and gives practical guidance, then reads it back to confirm the farmer understood.

If the issue is uncertain, matches a flagged high-risk category (like pest infestations that need visual confirmation), or the farmer explicitly asks for a person, the assistant stops guessing. It creates a structured case in the system, assigns a reference number, and tells the farmer their case is being routed to a human expert — preserving everything discussed so far.

Under the hood, Kisan Sahayak runs on Agora's Conversational AI Engine for real-time voice, speech-to-text, and text-to-speech, with an LLM handling natural conversation and a backend service matching symptoms against the advisory database and applying escalation logic. The result is a system that behaves less like a scripted bot and more like a knowledgeable first point of contact — one that knows what it knows, knows what it doesn't, and knows exactly when to hand off to a human.