Case study · Online poker · 2025
TablePoint
Online poker operator goes AI-native, table integrity first.
A 200-person poker operator handling 1M hands a day embedded AI across collusion detection, table management, and tournament ops — without eroding game integrity.
At a glance
- 200Headcount
- 4 weeksEngagement
- 8Pioneers
- 1MHands/day
- 24/7Tournaments
The brief
AI inside the integrity loop, not over it.
Cohort
- C-level (CEO, COO)2
- Heads of Department4
- AI pioneers8
Drawn from Game Integrity, Tournament Ops, Tech, and Customer Support — the four functions where player trust gets earned daily.
What the pioneers shipped
Five integrity-aware workflows.
Collusion-pattern detection
Surfaces multi-account and chip-dumping patterns; analyst signs every action.
Tournament structure copilot
Drafts blind structures, prize pools, and late-reg windows from operator briefs.
Bot/AI-assisted-play detection
Behavioural fingerprinting flags suspected bot accounts; integrity team reviews.
Tournament dispute summary
Assembles hand histories and chat into one digest for the dispute officer.
Player support categorization
Tier-1 questions auto-routed; integrity-related complaints escalated with full context.
“Integrity team has veto. Always.”
Approach
Four weeks. Integrity audit on every model.
Discover
Mapped poker stack, integrity controls, tournament ops. Defined what AI assists vs. what stays human-only.
Build
Pioneers built first cuts. Game-integrity team co-designed every detection model from day one.
Integrity audit
Three demos with Head of Game Integrity. Every detection model independently challenged with adversarial cases.
Compound
Skills shipped to internal marketplace; integrity team owns the lifecycle of every detection model.
Impact · Year one
Outcomes leadership signed off on.
Annual savings
€1.6M
Across the engagement's first twelve months.
FTE optimized
13
Full-time equivalents freed for higher-leverage work.
Business satisfaction
9.4 / 10
Post-engagement leadership review.
Why it didn't go off the rails
Tables are sacred. So are bans.
Integrity team has veto
Every detection-touching skill goes through integrity. Not product. Not BI. Integrity.
No auto-bans without human review
Agent flags. Integrity officer decides. Especially on cross-account and bot detection.
Transparent appeals
Every banned account gets the agent's reasoning attached to the appeal queue.
Audit trail per ban
Every action carries the agent log, the integrity-officer note, and the appeal record.
Where to start
Book an org assessment.
Ninety minutes, no slides.
We'll spend an hour with your leadership team mapping where AI creates the most leverage in your operation — and what an overlay would look like in practice.