A full-stack AI agent that decomposes queries into subquestions, runs tool-calling loops, and returns cited responses with real-time streaming.

An AI-powered Tinder-like sports recruiting platform that helps athletes find the best colleges and opportunities to showcase their talents.

The context layer for automotive repair. Turns fragmented OEM manuals into a single, queryable source of truth with page-level citations. Ask a question, get a verified answer.

A RAG-powered web application enabling users to upload files and retrieve contextual explanations using vector search and large language models.

A full-stack e-commerce platform with product listings, shopping cart, and checkout flow.

A web application providing Spotify users with deep insights into their listening patterns and music taste.

A ground-up implementation of the "Attention Is All You Need" paper, including multi-head self-attention, positional encoding, and full encoder-decoder architecture.

A convolutional neural network implemented to classify handwritten digits from the MNIST dataset with high accuracy.

An end-to-end machine learning project building a predictive system for California housing costs using real-world data pipelines.

A collection of machine learning projects covering regression, classification, clustering, and deep learning tasks.

Various OS-level projects including data and task parallelism, semaphores, and round-robin scheduling algorithms.