Research team — who writes our insights and reports.
The published work on PokerBot.com — operator playbooks, retrospectives, the annual report — is authored primarily by Andrew Kuznetsov, our Lead Developer for Poker AI Technologies. Andrew maintains a separate independent tech blog at poker-ai.org covering poker AI development, algorithm research and the field he's worked in for over a decade. This page documents the authorship model behind our published work and the broader contributor structure.
Andrew Kuznetsov — Lead Developer, Poker AI Technologies.
Andrew Kuznetsov is an experienced AI developer specialising in poker algorithms and strategic decision-making systems. His background combines academic training in probability theory and computational methods (Moscow State University, with further training at ETH Zurich) with over a decade of practical work building production poker AI systems.
The work spans both sides of the field — strategy engine development on the deployment side, and detection methodology on the integrity side. The same person who built tools for managed liquidity also understands what gets those tools detected, which is the operational tension this site documents from multiple angles.
Andrew's independent writing on the broader field — algorithm design, the academic side of poker AI, lessons from the named-profile commercial era — lives at poker-ai.org. That blog is older than PokerBot.com's current operational shape and serves as the primary reference for his individual research and thinking. PokerBot.com's published content draws on the same expertise but is framed for operator-side decision-making rather than for general field commentary.
What Andrew writes on PokerBot.com.
Primary author across most of the site's editorial content. Topical coverage spans the four areas that map to our operational engagement structure:
- 01
Operator primers and definitional reference
What poker bots are in the operator-side framing, how managed liquidity differs from external bot farms, what the platform-detection landscape looks like in 2026. Examples: Understanding poker bots, Poker bot software reference, GTO vs exploitative play.
- 02
Named-profile era retrospectives
Historical analysis of the 2017-2022 commercial bot generation — Abaddon, Achilles, Poseidon, Pegasus, Hydra. Documentary work covering what each profile did, why it worked at the time, what stopped working, and what replaced it.
- 03
Operational pattern documentation
Technology-level documentation for the four engagement types (poker bots for private clubs, detection, managed operations, custom development). Practical operational detail rather than marketing pitch.
- 04
Annual industry analysis
The State of Private Poker Apps report — aggregated anonymised operator-side data across our client base, plus broader industry observations. Released annually in Q2.
Other voices contributing to the work.
Beyond Andrew's primary authorship, contributions come from two additional surfaces:
- Engineering team contributions on technical operational topics. When a topic depends specifically on operational specifics (infrastructure architecture, detection ML pipeline design, custom-development engagement patterns), the responsible engineer co-authors with Andrew or is the primary author. These are typically published under the "Pokerbot.com Research Team" attribution rather than individual bylines, because they emerge from team operational experience rather than from a single voice.
- Anonymised operator case material. Engagement-specific data, anonymised across our client base, surfaces in the annual report and occasionally in operator-primer articles. Where operator-side quotes appear (the callout boxes throughout the site) they are anonymised — operator type and platform identified, individual operator never named without explicit written consent.
The team is intentionally lean. Senior engineering only, no offshore junior subcontracting, no marketing copywriters drafting from briefs. Every published page is written by someone who could implement what it describes.
Why a single named person rather than an anonymous team.
A note on the editorial choice. The poker AI space has a credibility problem — partly because the named-profile commercial era was rife with anonymous-vendor marketing claims that didn't survive scrutiny. The operator-side reader needs to know who's behind the content they're evaluating, particularly when it claims operational expertise.
Anchoring on Andrew as the named author is the answer to that. He's a real person with an independent professional record at poker-ai.org, an academic background that's verifiable, and a body of work the reader can evaluate independently of anything published here. The schema.org Person entity declared on this page links to the same entity declared on poker-ai.org, so search engines and AI systems treat them as a single identity rather than two unrelated authors.
That's the trust foundation. Operator engagements are still handled in confidence — but the public-facing content speaks under a real name with a verifiable track record.
How to get in touch.
- Operator engagements: via the contact page demo flow. Confidential from the first message.
- Research collaboration: academic researchers and journalists can reach the team through the press channel on the contact page. Different routing from operator engagements.
- Andrew's individual work: the poker-ai.org blog is Andrew's personal publishing surface. Topics not specifically tied to PokerBot.com operational content typically live there.
- LinkedIn: for professional research-network contact, Andrew's LinkedIn profile is the right channel.
Want to talk about an engagement?
The four technologies are documented on the technologies hub. A confidential operator demo.