At Social Prachar, I clean and analyze tourism, climate, and technical-support data, then use it to surface trends and prepare model-ready datasets. I built XGBoost models for trip-cost prediction and crowd forecasting, along with a PyTorch LSTM model for climate forecasting.
I also developed an NLP pipeline that classifies support issues and retrieves relevant troubleshooting guides. In my Telangana Tourism Trails project, I brought cost, crowd, and climate insights together in a trip-planning app using Streamlit, FastAPI, Supabase, and SQLite.
