{"slug":"forexery","name":"Forexery","hackathon":"hackprinceton-fall-2025","tagline":"A reinforcement-learning agent that trades USD/JPY using news sentiment plus technical signals.","placement":{"status":"track-winner","awards":["Best Predictive Intelligence by Chestnut Forty"],"track":null},"team":{"size":2,"members":[]},"problem":"Most retail forex traders lose money; technical indicators miss why markets move.","solution":"Trains an RL agent in a trading environment combining sentiment from news with technical features.","demo_moment":"Team-described: compare the agent's returns against buy-and-hold.","build_style":"software-only","tech":{"stack":["python","gymnasium","pandas","numpy","tensorboard","dedalus"],"ai_patterns":["ml-model","llm-api"]},"domains":["fintech"],"analysis":{"why_it_stood_out":[{"claim":"Won “Best Predictive Intelligence by Chestnut Forty” among 195 submissions.","basis":"organizer-stated","source":1},{"claim":"Combines why (news) and what (price) in one model.","basis":"team-stated","source":0}],"mechanisms":["visible-engineering-depth"],"judging_lens":["technical-depth"],"lessons":["Show training curves and a baseline when claiming trading results."],"confidence":"low"},"links":{"devpost":"https://devpost.com/software/forexery","repo":"https://github.com/codejvn/forexery","demo":null,"video":null},"sources":[{"url":"https://devpost.com/software/forexery","kind":"devpost-project","retrieved_at":"2026-10-13"},{"url":"https://hackprinceton-fall-2025.devpost.com/project-gallery","kind":"devpost-gallery","retrieved_at":"2026-10-13"}]}
