Kishan (কৃষাণ) AI Platform
National agricultural intelligence: daily DAM market prices, hyper-local weather alerts & Bengali voice assistant.

~/projects/kishan $ glow README.md
Why I built it
Farmers across Bangladesh lose hard-earned income to price manipulation by middlemen and unpredictable flash weather events. Kishan was engineered to deliver instant price transparency, AI agronomic guidance, and voice-assisted weather alerts directly into the hands of farmers.
What it does
- Tracks daily wholesale vs. retail produce rates from the Department of Agricultural Marketing (DAM) with clear trend indicators (📈/📉).
- Generates hyper-local 7-day weather forecasts with crop-specific "What to do" and "What to avoid" action plans.
- In-app Bengali Text-to-Speech (TTS) audio engine reading prices and forecasts aloud with one tap.
- Pluggable SMS emergency gateway for early warning disaster and heavy precipitation alerts.
- Offline-first local caching and passwordless guest onboarding for zero friction.
How it's built
- Full-stack architecture: Flutter Android mobile application paired with a high-throughput Python FastAPI backend.
- AI-assisted market analysis with Google Gemini API and automated web scrapers.
- Open-Meteo meteorological API integration and SQLite cache layer.
"dependencies": ["Flutter", "Dart", "Python", "FastAPI", "Gemini API", "Open-Meteo", "SQLite"]




