About PokerBot.com — operator infrastructure, not consumer software.

Four technologies, one operator-aligned playbook.

The work we do for clients falls into four technology categories — covered in operational detail on the Technologies hub. Brief framing here:

1. 01

   Poker bots for private clubs

   Operator-deployed AI seats configured for break-even balance — keeping tables alive during off-peak windows without extracting from real players. Our flagship engagement and the entry point for most club owners.

2. 02

   Poker bot detection

   A detection overlay that surfaces external bot farms, RTA and collusion invading the operator's club. Behavioral biometrics, statistical audit, collusion graphs, network fingerprinting. Bundled with liquidity or run standalone.

3. 03

   Managed operations

   Fully-managed operations for partners who have access to active clubs but no ops team. We run the whole operational lifecycle — deployment, account management, monitoring, footprint discipline, settlement — you provide access.

4. 04

   Custom development

   Bespoke engagements for problems outside the standard grid. Platform integration, custom strategy engines, operator-owned detection models, hand-history and data-engineering pipelines.

Operator-side audience, specifically.

Our entire engagement model is built around private-club operators — unions, club owners, club operations leads. A few specifics to anticipate common questions:

• Private-club platforms only. Active operator relationships across PPPoker, ClubGG, PokerBROS, Suprema, HHPoker and X-Poker. We don't operate against PokerStars, GGPoker, WSOP.com, partypoker, or any state-licensed public operator.
• Union-scale operations. Most of our engagements serve unions with 10-500 clubs and 500-50,000 monthly active players. Smaller single-club operators are sometimes a fit; very large unions (5,000+ clubs) typically need custom engagement scoping.
• Multi-region geography. Operator base centred on Asia/global private-club economies but extending across regions where the platform ecosystem has reach. Engagements are jurisdiction-agnostic from our side; operator-side compliance with local regulation remains the operator's responsibility.
• Not individual players. Consumer-side poker software for individual play is handled by a separate operation entirely — different brand, different audience, different product. Not relevant to operators reading this page.

The four operational commitments.

Four principles that govern how engagements run. These aren't marketing values — they're commitments written into every contract.

1. 01

   Confidential from the first conversation

   Operator identity, deployment specifics, hand-history data, financial details — all kept in confidence from the first message. No public case studies with names. No screenshots of client clubs. No referrals revealed without explicit written consent.

2. 02

   Break-even ecology, not extraction

   Managed Liquidity deployments target aggregate monthly P&L within ±3% of zero across all AI seats. The operator profits from rake on hands played; AI seats are configured for presence, not for winning money from real players. This is the only defensible long-term position and we don't compromise on it.

3. 03

   Operator-controlled policy

   Every engagement leaves the operator with full credentialed control. Stakes, hours, behaviour profiles, seat counts — all set by the operator and reversible from a control panel. Nothing runs on vendor defaults the operator can't override.

4. 04

   Confidential single-tenant infrastructure

   Each operator's deployment runs on isolated infrastructure. No pooled tenancy, no shared binary distribution, no cross-client configuration visibility. Detection patterns don't leak across operators because deployments are configured independently.

What we explicitly don't do.

The poker-automation industry has earned a credibility problem over two decades. We address it the only way that works long-term: by being explicit about scope.

• We don't sell consumer-facing player bots. Software designed to be sold to individual players for personal-play use is a separate operation entirely — different audience, different product. We don't blur the two lines, and it isn't handled on this site.
• We don't extract from real players. Managed-liquidity configurations target break-even economics. Bots that win money from real players are a different category we don't build, sell, or operate.
• We don't disclose client identity to other clients. Every engagement is kept confidential. No case studies with names. No referrals revealed without written consent. Client base remains opaque to the public — and to other clients on our books.
• We don't run black-box deployments. Operators see everything we see — same dashboards, same metrics, same alerts. If we can't be transparent with the operator about what's happening in their own club, we shouldn't be running the engagement.

What we publish, and who writes it.

Beyond the operational technology work, we publish research and reference material in three categories:

• Operator playbooks and tutorials. The tutorials archive covers the named-profile era retrospectively and the modern operator pattern forward-looking. Educational rather than commercial.
• Insights and reference articles. The insights section covers individual topics — GTO vs exploitative play, bot detection, the poker economy of private clubs — with operator-side framing.

Most published work is authored by Andrew Kuznetsov, who maintains a separate tech blog at poker-ai.org covering poker AI development and adjacent topics. The research team page covers the broader publication structure and contributor model.

How the company actually operates.

Brief operational framing — senior-led engineering, operator-side focus:

• Hong Kong-based operating entity serving an international operator base. The jurisdictional choice reflects the regional concentration of private-club platform operators.
• Senior-led engineering team. Custom development engagements are handled by the same people who scope them, so operators talk to the engineers doing the work rather than an account layer in between.
• Engagement model built around fit. We sometimes decline engagements when the operational discipline cost is too high or when the fit isn't right — the work is selective by design, not transactional.
• On the domain since 2002, with long-standing operator relationships. The current operational shape — the four technologies described above — took its present form over the last several years, after the named-profile commercial era gave way to the managed-engagement model.