AI Krishi-Saathi: An Agentic AI Assistant for Farmers in India
Small and marginal farmers face major challenges: late crop disease detection, unpredictable weather, and unfair mandi pricing. Existing advisory tools are often English-only, app-based, and inaccessible to rural farmers with basic phones.
AI Krishi-Saathi is a Generative AI-powered assistant, available via WhatsApp and IVR calls, that delivers personalized guidance in local languages. It uses a multi-agent system:
Agent 1 – understands queries and translates local dialects.
Agent 2 – diagnoses crop issues and suggests remedies.
Agent 3 – fetches real-time weather, mandi prices, and government scheme data.
This simple design allows farmers to send voice or text queries and receive accurate, localized, and actionable advice directly on their phones.
By combining accessible communication channels with AI, Krishi-Saathi can empower 120M+ farmers, reduce crop losses, and increase incomes by 10–15%. The system is inclusive (supports 20+ languages), scalable, and aligned with the IndiaAI Mission to apply AI for social impact.
With phased implementation—starting as a WhatsApp bot, integrating IBM ADK, and piloting in one district—Krishi-Saathi is a feasible, high-impact solution for Indian agriculture.
Built With
- agmarknet-(mandi-prices)-ai-tools:-ibm-adk
- hugging-face-transformers-apis:-twilio-(whatsapp)
- language:-python-frameworks:-langchain
- open-meteo-(weather)
- streamlit
- whisper-api
Log in or sign up for Devpost to join the conversation.