Projects
Merge Fitness
A companion mobile app for Merge Fitness LLC (120+ users). Allows trainers to assign workouts/nutrition plans to clients, track their progress, and schedule sessions. Supports live trainer location tracking, video sessions, and in-app payments.
Resume Hub
Resume search tool for GT Business Fraternity (110+ users) that parses resumes with AI and allows users to find those with similar major, year, skills, and experience. Supports batch uploading of 100+ resumes at once.
Interview Pro
A web application (used by 50+ college students) where users can practice their interview skills and receive detailed feedback on their content, structure, and body language using LLMs and OpenCV + Tensorflow.
Trading Platform
A trading platform UI for Trading at Georgia Tech trading competitions (400+ users), integrated with a matching engine that facilitates order management and real-time price updates. Provides tools to interact with the matching engine by submitting orders, monitoring trades, visualizing the order book, and tracking market prices.
Additional Projects
C++ High Frequency Market Maker
Deployed latency-optimized High Frequency Trading market maker with Coinbase level-2 order book achieving sub-0.1ms order-to-fill execution. Built multi-threaded system with atomic ops, lockless structures, Redis caching, & thread-safe execution at 120+ trades/sec. Devised real-time microsecond risk management engine preventing 99.9% of drawdown violations across 50M+ simulated trades.
CC++RedisLibWebSocketsCoinbase APIAWS EC2Alpha Kappa Psi Rush Application
Web application for GT's Alpha Kappa Psi fraternity to centralize information of 200+ potential new members/rushees. Used by 100+ fraternity members to expedite the rush process through a fully digital collaborative platform.
SpeakEasy
A web application that allows users to practice their public speaking skills in a variety of scenarios (recruiters, significant other's parents, etc.) by recording and analyzing their speech and offering LLM-based feedback to improve their performance. Early precursor to Interview Pro project.
MediSense AI
Engineered a full-stack web application that helps patients easily find diagnoses for conditions based on symptoms. Tuned Google BERT LLM predicting prognoses with 91% accuracy across 150+ trials with 50,000+ medical reports. Trained Logistic Regression & Random Forest models assessing risk factors with 83% accuracy on 560,000+ rows.
Premier League Betting Assistant
Created dashboard to help Premier League bettors make more informed decisions. Applied Selenium to webscrape data from DraftKings, Bet365, and Pinnacle to compare real-time PL match odds and find arbitrage opportunities. Calculated the percentage chance of bets to hit with scraped data and used D3.js to visualize the data.
PythonSeleniumD3.jsNumPyPandasMovies Store Application
Website that allows users to view, explore, and buy movies while creating and reading user reviews. Project for CS 2340 Objects and Design class.
Personal Portfolio
This very website you are on right now.