Smart-Agro
About the project
SmartAgro is a multilingual, AI-powered advisory web app (PWA) built for marginal Indian farmers.
It targets the main barriers farmers face:
- Most agricultural advice is in English rather than regional languages.
2.Crop diseases are spotted late because experts aren’t nearby
3.Mandi prices are scattered across slow government portals and quoted per quintal when small sellers think in kg.
4.Raw weather forecasts don’t say what action to take, and low literacy, weak networks and low-end phones make heavy apps hard to use.
To address these, it offers an
- AI chatbot that answers farming, MSP and Kisan scheme questions in the user’s language,
- Using live weather and price data when needed.
- It also offers crop disease diagnosis from a crop photo, giving the likely disease, severity and treatment advice. A pre-check rejects non-crop images, and the result shows which model answered.
- A Mandi Prices page shows live Agmarknet data with a quintal/kg toggle and falls back to cached prices when the government API is down. 5. Weather forecasts are turned into a 15-day risk outlook, with seasonal advisories and humidity-based pest and fungal warnings.
- An NDVI gauge shows crop health, honestly labelled Satellite.
- Voice input and text-to-speech support low-literacy users. Settings cover 23 Indian languages,
- light/dark theme, °C/°F and price units, and the installable PWA works offline with cached pages.
Built with
geminipythonjavascriptflaskhtml5geminiapigemmagroqgithubcss3vscodeapijsonllamarender
AI tools
Coded withClaude CodeCodexAntigravityGitHub CopilotBoltGemini 3.5 Flash LiteGemma 4 31BClaude Sonnet 5.5GPT-6 LunaQwen3.8 Flash
Prize Categories Entered
- Submitted
Best Use of Gemma 4
- Submitted

Best Open-Source AI Project