B U

I built an E-commerce Support Resolution Agent with a RAG pipeline using LangChain, Gemini API, Hugging Face embeddings, and FAISS. It retrieves policy text, generates replies grounded in that text, and rejects unsupported responses; I tested it on tricky tickets and refined the prompt based on the results.

For my Smart Email Summarizer, I developed a DistilBART-based application with a FastAPI backend and Chrome Extension. I tested different inputs and text lengths to find and fix inaccurate, incomplete, or irrelevant outputs.

I also built a Flask application for heart disease risk prediction, compared Logistic Regression, SVM, and XGBoost, and selected XGBoost as the final model. My Battery-Aware EV Route Optimization System used distance-based energy estimation and charging-feasibility logic; I published related research in IJISRT in 2026.

During my internships, I created a Tableau dashboard from JSON telemetry data at Deloitte to analyze machine downtime, and developed a responsive music player at Coding Rajas. I’m a 2026 B.Tech IT graduate seeking an AI internship where I can assist in building, testing, and documenting AI/ML solutions.