Introducing GreenLight: an exceptional native iOS application meticulously crafted by a classmate and me. This innovative app harnesses the power of OCR technology to seamlessly scan Dartmouth Student IDs, revolutionizing the way we manage event entry and enhance security protocols. With "Greenlight," event organizers can effortlessly track guest entrances while simultaneously enforcing security measures, including real-time verification against a comprehensive event blacklist. Won second place in the Empowerment category at HackDartmouth VII in April 2022.
Presenting a Python-based artificial intelligence project that indulges my passion for chess: a formidable AI designed to take on opponents on the chessboard. Harnessing the potent combination of Minimax search and fortified with Alpha-Beta pruning, transposition tables, and judicious lazy evaluation, this AI is a true strategic marvel. This project is on-going. My current focus is on enhancing performance by parallelizing the search, and I'm diligently assembling a comprehensive library of openings and endgame strategies. As it stands, this AI boasts an approximate ELO rating of 1500. Stay tuned for exciting developments.
A tiny search engine for Unix CLI featuring a web page crawler, indexer, and querier, complete with custom set and hash table modules for efficient data management. By utilizing Boolean logic for parsing and integrating web scraping capabilities to retrieve linked URLs, I provided ranked search results for user queries.
I employed NumPy and Python to create an artificial neural network entirely from scratch. This project was a deep dive into neural network architecture and training. The network includes alternating linear and non-linear activation layers followed by a loss function. The model's parameters are fine-tuned using gradient descent.