Making large documents queryable and explainable using LLMs.

• •Designed a LangGraph-based reasoning workflow to dynamically route user queries based on intent.
• •Integrated vector-based semantic search for efficient context retrieval.
• •Improved response relevance by selecting task-specific reasoning paths during inference.

Building a production-style full-stack platform.

• •Implemented secure authentication workflows and complex MongoDB schemas.
• •Designed scaleable MVC architecture with modular routing and session management.
• •Enabled dynamic server-side rendering using EJS and Express.js.

Predicting health insurance costs using ML.

• •Built separate ML models to predict premiums based on demographic data.
• •Optimized XGBoost hyperparameters for high accuracy prediction.
• •Deployed a real-time interactive Streamlit dashboard for end-users.