Prajit Chennupati is a Computer Science student at the Georgia Institute of Technology’s College of Computing, specializing in Intelligence and Systems and Architecture, with a minor in Mathematics. His academic interests span artificial intelligence, computer systems, algorithms, and mathematical foundations of computing, with an emphasis on developing efficient, reliable, and scalable computational systems.
Prajit has applied these interests as a developer at XR DOJO, where he contributed to ATL Quest, a World Cup-themed augmented reality soccer game developed in anticipation of the 2026 FIFA World Cup. Using Unity, C#, and AR Foundation, he implemented interactive gameplay mechanics, including AR tracking and ball physics. He also optimized runtime performance by eliminating heap allocations in performance-critical code paths and reducing redundant per-frame updates, cutting CPU frame time by 30%.
Beyond his work in augmented reality, Prajit developed an MLB forecasting platform that applies machine learning to baseball prediction. Built with Python, FastAPI, Next.js, XGBoost, and PostgreSQL, the platform evaluates historical seasons and generates probabilistic forecasts, achieving a log loss of 0.683 compared with 0.693 for a coin-flip baseline. This work reflects his broader interest in combining rigorous technical foundations with practical applications of machine learning and software engineering.
Essays
Projects
- Forecasting the 2026 MLB Season
A preseason forecast of every 2026 game, the standings and the postseason, built only from data through 2025 and graded against what happened.