The fastest way to improve at chess is to practice the exact patterns you keep missing — not random puzzles from a corpus. Chess Masti AI turns your own blunders into puzzle drills automatically: lose to a knight fork once, practice five fork positions next. The training loop is what moves rating points; pattern recognition is built one mistake at a time.
Standard puzzle sites give you a random sequence — a back-rank mate from someone else's game, a Russian endgame from 1985. The patterns are real but they aren't your patterns. Your blunders cluster: most players lose to the same handful of mistakes over and over.
Drilling random tactics is fine. Drilling your own tactics is faster. The pattern your brain just failed to see in a real game is the exact pattern most worth reinforcing while it's fresh.
Find the moment
Stockfish evaluates your game. The biggest evaluation swings are the moments worth training — blunders that lost material or missed a winning tactic.
Convert the position
Chess Masti takes the position from one move before your mistake. Now it's a puzzle: the engine's best move is the solution. You see it before you played; now you play it again with the answer not yet visible.
Drill it
Solve the puzzle. If you got it right this time, the pattern is one repetition stronger. If you got it wrong, try again — the system shows the solution and explains why.
Pattern stacks
Over a few games, the same kinds of mistakes start clustering. Knight forks. Back-rank weakness. Passed pawn miscalculation. The drill set self-organizes around your real weaknesses.
This is not a replacement for tactics training on a curated set. Chess Masti's drill set is built from your real games — if you only play 10 games a month, you only have 10 games' worth of mistakes to drill. That's often enough for the patterns to accumulate, but for raw volume, pair it with a tactics trainer (Lichess puzzles, Chessable, etc.).
This is also not a spaced-repetition system in the strict Anki sense. The drills are available for repeat practice, but interval-based scheduling is on the roadmap, not in the product today.
See the full pipeline
The free AI chess coach page covers how the engine + AI + mistake-detection + puzzle generation all fit together.
Open the free AI chess coach pageDrill your own mistakes
Paste a PGN. Find your blunder. Practice the exact pattern. Free.