Name the idea
Reconstruct what you were trying to achieve instead of treating the move as an isolated mistake.
Post-game Go learning
Understand why your Go move missed the point.
Go AI Coach is a post-game learning app for Go, Baduk and Weiqi. It is being developed for adult players—especially DDK and SDK improvers—who want more than a number beside a move.
In development. Go AI Coach is not yet available as a public service.
After the game
Your idea
“I wanted to protect the corner.”
Start with the intention behind the move.
Position evidence
What was urgent here?
Compare the idea with the whole-board situation.
Practice
Recognise the decision again.
Turn one review into a reusable pattern.
From score to lesson
Analysis should help you think.
A move can lose points without revealing what the player misunderstood. The intended Go AI Coach experience begins with the reason, then connects it to verified position evidence and a practical next step.
Test it against the board
Separate factual position evidence from coaching language, so explanations remain grounded.
Make it reusable
Connect recurring decisions to patterns and future drills, rather than ending at a verdict.
Intended workflow
A review built around your reasoning.
The product is being designed for completed games only. The sequence below describes the intended experience, not features currently offered as a public service.
Finish the game
Review begins after play, never as live-game assistance.
Replay the thought
Reasoning Replay helps capture what you noticed and intended.
Compare
Place your idea beside the position’s concrete demands.
Ask why not
Explore why a natural-looking move may fail to answer the real problem.
Practise again
Use recurring patterns and drills to reinforce the decision.
Designed for explanation
Two questions at the centre.
Raw analysis can identify a better line. Coaching should also help a learner understand the decision that led away from it.
What were you trying to do?
The intended experience reconstructs the player’s idea before explaining the position. That gives the lesson a human starting point without presenting AI language as board truth.
Why did the natural move miss?
The goal is to compare plausible alternatives, show what changed, and make close decisions sound close—without overstating certainty.
Our approach
Evidence first. Language second.
Go AI Coach is being designed so deterministic game facts remain authoritative, while generated language is limited to explaining validated evidence.
See how the layers fit togetherEvidence-backed lessonsPosition facts, candidate comparisons and lesson claims should remain traceable.
Honest uncertaintyClose calls should not be turned into dramatic or absolute conclusions.
Patterns and drillsA review should reveal recurring decisions and support deliberate practice.
Post-game onlyThe product is not intended for assistance during an active game.
Who it is for
For players learning to see more.
The initial product thesis focuses on adult improvers who can read an analysis result but still need help converting it into a practical Go lesson.
Especially relevant to
- DDK players building whole-board awareness
- SDK players refining judgement and priorities
- Players reviewing Go, Baduk or Weiqi after a completed game
Not positioned as
- A tool for live-game help
- A replacement for a human teacher
- A promise of guaranteed rank improvement
- An official partner of a Go federation or platform
Development status
Building carefully, before inviting players in.
The application is in pre-alpha development. There is no public account, purchase, game-review service or release date on this website today. This page will be updated when those facts change.
Pre-alpha
2026
2026