CS (Data Science) student, final year — building at the intersection of data science and backend engineering.
AI-powered flood relief decision support system for Kerala, built on real 2018 flood data, Census population records, and historical rainfall data.
- Rigorously evaluated ML models via Leave-One-Out Cross-Validation; pivoted to a transparent, interpretable scoring approach when no model beat a baseline — full reasoning in 8 documented Architecture Decision Records
- Priority-weighted resource allocation via linear programming (PuLP)
- Full-stack deployment: FastAPI, PostgreSQL, Docker, Streamlit
Python · FastAPI · PostgreSQL · Docker · SQLAlchemy · Streamlit · scikit-learn · XGBoost
[LinkedIn - https://www.linkedin.com/in/aparna-singh-150444255] · [Email - aparnasingh081003@gmail.com]