Poker AI lab. Coach product.
Opponent modeling from 6.6 million analyzed decisions — wired into exploit reads, leak drills, and pressure-test sim. Not a spot pack. Not raw solver output.
Vision — research in progress
Unshipped R&D: live player tells, gaze, and table-state mapping on broadcast footage. Not a shipped product feature — a lab project exploring the layer underneath solver lines.
// $687K pot · Tom Dwan vs Wesley · tells logged in real time · tap a frame to jump
Baseline-locked reads
Every player gets a neutral baseline from their own face at rest. Stress, cognitive load, and control are measured as deviation — not vibes.
Named tells, not noise
Lip compression, nostril flare, gaze lock, eyebrow flash — flagged when they spike above that player's normal range.
Scene map on the felt
Chips, cards, and players detected live. Pot and board synced to the broadcast. We measure — we don't guess.
Gaze + pressure overlay
Where are they looking — opponent, board, away? Pressure and deception risk bars update as the hand escalates.
Poka is building the layer underneath — who's under pressure, who's composing a story, and whether the line matches the read.
Exploit reads grounded in opponent modeling.
We analyzed how 36 frontier models play poker and built archetype models — the coach tells you how to attack stations, nits, and regs in plain English, with your measured leaks when you have data.
- +Archetype exploit reads — not generic ChatGPT
- +6,000+ curated strategy sources behind every answer
- +Personalized leak reads with real accuracy % and sample sizes
- +Always-on dock + full chat at /ask
> ask
How do I beat a calling station?
> coach
Value bet thinner and bigger — they pay off dominated pairs and draws. Cut bluffs unless you hold strong blockers; stations rarely fold. Size for extraction, not balance.
See it on your own game.
Seven days, full access, no card. Your leak profile starts building from drill one.
Start free trial →