Kisan AI
WhatsApp Assistant for Indian Agriculture
WhatsApp-first multilingual AI assistant providing personalized crop diagnostics, weather alerts, and mandi prices for Indian farmers.
The Engineering Challenge
Smallholder farmers in regional India lack real-time personalized agronomic guidance in their native dialects.
Core ProblemDeliver actionable agricultural insights via low-bandwidth WhatsApp interface with voice note support.
- Regional language nuances (Hindi, Malvi, Marathi)
- Low-bandwidth image and audio compression
System Architecture
End-to-end data pipelines and component isolation boundaries.
Twilio/WhatsApp Business webhook pipeline routing audio queries to Whisper multilingual STT, LangGraph agronomy advisor, and regional weather APIs.
Key Engineering Decisions
Voice-First Multilingual Interaction
Engineered voice note ingestion so non-literate farmers can speak crop symptoms naturally.
Production Verification & Results
- Currently in active development with field validation trials
- Prototype processes regional voice queries and returns synthesized audio guidance
Production Stack & Tooling
Want to build something similar?
I specialize in taking complex agentic workflows, duplex voice architectures, and high-throughput computer vision models into production.