Projects
Production-style tools — not just notebooks — spanning MLOps, medical image segmentation, and LLM-powered applications.
End-to-end retinal vessel segmentation platform built on an Attention U-Net, with a Streamlit UI, MLflow tracking, and Docker-based deployment.
Production-ready MLOps platform for model deployment, benchmarking, inference, latency analysis, experiment tracking, and REST API serving.
AI-powered agentic assistant — a Flask-based LLM application with RAG, memory, document analysis, and tool integration.
RAG document intelligence system for document-based question answering with persistent vector search.