Projects
A selection of projects I have worked on.
ReadmeRag — RAG Chatbot for GitHub Projects
A lightweight, zero-cost RAG (Retrieval-Augmented Generation) chatbot that answers questions about a developer's GitHub projects. It fetches README files from a GitHub profile, indexes them in a local vector database, and uses a free-tier LLM to answer questions.
End-to-End GCP Big Data Healthcare Pipeline
Led a team of 5 to build a GCP data lifecycle framework for chronic disease prediction, routing raw medical data through BigQuery to Looker Studio.
Agentic Medical AI Pipeline
Web app using LangGraph state machines for medical data visualization and imputation, powered by Azure GPT-4 and local Qwen2.5.
Google Cloud Data Management Project
Led a team of 8 to architect an end-to-end GCP pipeline using PySpark and Hive on Dataproc to process hotel datasets for BI insights in Data Studio.
Low-Light Object Detection Pipeline WebApp
My capstone project: Web app with a sequential preprocessing pipeline (denoising + enhancement) and custom-tuned YOLO, achieving SOTA mAP50 on the ExDark dataset.
DermAI — Skin Lesion Classification
Flask-based web app for classifying skin lesions using ML models. Supports CNN, EfficientNet, and CNN+Metadata models. Upload images, choose models, and view predictions with confidence scores.