“My project is called AnnData. It is a mandi price advisory app for farmers in Madhya Pradesh.

The problem: Farmers often don’t know which nearby mandi pays the best price for their crop, so they sell at a lower rate.

What I built: The app takes mandi price data from the government’s data.gov.in API and cleans it with pandas. Then a FastAPI backend compares prices across mandis and tells the farmer the best mandi for their crop and how much extra they can earn per quintal. The frontend is a Streamlit dashboard with Plotly charts, and I deployed it on Render.

What I learned: Government data is messy and has missing values. I learned how to fetch data from an API, clean it, and turn it into a simple, useful result for a farmer.

Next step: Right now the price trend chart is a simulation. I plan to train a real forecasting model on historical mandi data.”