Geospatial AI · Agriculture
Crop Recommendation Engine
Soil, weather, and market signals brought together to help farmers compare crop choices for a specific location.
The problem
A crop can suit the soil yet perform poorly against weather or market conditions. Useful recommendations need to consider the farm’s location, soil profile, growing period, and access to nearby markets together.
Features
- Location-based soil and environmental dashboards
- OCR-based nitrogen, phosphorus, and potassium extraction from soil reports
- Historical weather trends and LSTM forecasting
- Crop-price forecasts and nearby-market context for recommendations
My contribution
Developed a geospatial dashboard and crop-recommendation pipeline combining soil-report extraction, satellite-derived parameters, weather trends, and crop-price forecasts.
The approach
01
Read the land
Mapped farm coordinates into geospatial raster datasets to retrieve soil and environmental parameters. The dashboard combines these with soil-report values and current weather.
02
Look beyond current conditions
Used historical trends and LSTM forecasting to examine conditions over a growing period. Market analysis considers harvesting timelines and proximity to APMC markets.
03
Bring the signals together
A Gemini agent combines GIS, weather, soil, and market information into crop recommendations. The React and Plotly interface makes the underlying signals explorable.
The outcome
A grand-finalist project at IIIT Lucknow HackoFiesta 6.0, combining farm conditions and market context in an interactive decision-support prototype.
Next exploration