I’m a high school player building an AI chess coach that finds why you blunder, not just what. Need beta testers!
Hi everybody, I’m a junior in high school and a competitive varsity player. Like a lot of you, I got tired of staring at the Chess.com game review screen and not improving from it. Seeing a red bubble that says "Blunder" tells you what went wrong in that exact moment, but it doesn't tell you why you keep making that same class of mistake game after game.
To fix this, I’ve been building ChessDx. Instead of move-by-move analysis, it’s an automated pattern diagnosis engine.
Here’s how it works:
- Game Ingestion: It automatically pulls your last 50–100 games via the Chess.com or Lichess public APIs.
- The Diagnostic Core: It runs your history through Stockfish, calculates centipawn loss, and maps your blunders across 6 dimensions: Openings, Tactical Patterns, Positional Play, Endgame Technique, Time Management, and Psychological/Tilt.
- The Diagnosis Profile: You get a clean radar chart showing your exact breakdown of mistakes.
- Adaptive Training Plans: An agentic layer takes your profile and builds a structured, 7-day training plan with daily puzzle sets pulled dynamically from the Lichess database to target your specific bottleneck.
I'm also building out a team-tier dashboard specifically for high school chess coaches so they can track an entire roster's weaknesses, assign drills, and spot regression instantly.
I’m opening up a private beta soon to test the data pipelines and the classification accuracy before the next school season hits. If you're a club player (800–1600 ELO) or a team coach, I’d love your feedback.
If you want to grab a spot in the beta or just help me out by answering a few quick questions, please fill out the form below!