Poker Coach AI is a specialized poker coaching agent that analyzes hands and trains range-based decisions across No-Limit Hold'em, Pot-Limit Omaha, cash games, and MTTs. Unlike generic chatbots that forget leak history, this agent retains your player profile, Active Leaks, and opponent notes. Built on Jenova's unified platform, it combines models from OpenAI, Anthropic, Google, and more with hand-history parsing and study tools. Poker has variance — advice is framed as +EV decisions, not guaranteed results.
Poker Coach parses standard hand history exports from PokerStars, GGPoker, Winamax, 888, WPT Global, and similar formats, and also accepts natural-language hand descriptions. When positions, stack depths, or reads are missing, the agent asks clarifying questions instead of filling gaps with guesses. When several hands are shared, analysis looks for cross-hand patterns — recurring tendencies, frequency imbalances, and systematic leaks — before treating any single river as the whole story.
Poker Coach treats GTO-informed play as the default backbone, then layers exploitative adjustments when you describe a specific regular, a HUD read, or a population tendency at your stake. At micro and low stakes, heavy exploitation is treated as the higher-EV path; at high stakes, balanced ranges and minimum defense frequency are treated as survival. Recommendations name combos, specify position and stack depth, and state the tension between the theoretically correct line and the practically profitable one.
Poker Coach runs single-decision and multi-street drills with full context — effective stacks, positions, prior action, board texture, and relevant reads — then waits for your action before revealing analysis. Level is inferred from how you talk: pot odds and tight/loose language versus ranges, combinatorics, polarization, node-locking, and ICM. Observed weaknesses, such as turn sizing in 3-bet pots or the river bluff-catch/fold boundary, become the next drill focus rather than a generic difficulty slider.
Poker Coach addresses tilt before returning to strategy, naming types such as entitlement tilt, revenge tilt, injustice tilt, winner's tilt, and desperation tilt, and calling out biases like results-oriented thinking, confirmation bias in reads, anchoring to initial hand strength, and loss aversion. Session structure covers pre-session routines, stop-loss and stop-win rules, time-boxing, energy management, and game selection. Bankroll management stays format-specific — cash versus MTT buy-in rules, moving-up criteria, shot-taking protocols — and is coaching, not investment advice.
Poker Coach treats solver work as a way to understand why an action is chosen, not a frequency-memorization drill, and ranks study as highest-frequency spots first, then highest-stakes decisions, then edge cases. Range charts can be generated as CSV files, study plans and session templates as PDFs, coaching recaps as DOCX, and board-texture or range-grid visuals as images. For current tournament schedules, rake structures, platform changes, or training-site offerings, the agent searches rather than inventing listings.
| Traditional Approach | Poker Coach |
|---|---|
| Forgets leaks after one session | Tracks Active Leaks across sessions |
| Recites GTO without stake context | Layers exploits onto GTO backbone |
| Vague advice missing position and stacks | Specifies position, depth, and pot odds |
| Can't parse exported hand histories | Parses PokerStars and GGPoker exports |
Unlimited Memory vs. Chatbot Amnesia A generic chatbot that drops context mid-coaching will re-diagnose the same river overcall, lose notes on a regular who never bluffs after a check-call-check line, and reset you from ICM language back to beginner pot-odds analogies. Poker Coach retains format, stakes, environment, platform, coaching mode, Active Leaks phrased as observable tendencies, opponent notes, and drill-performance patterns. Resolved leaks stay marked so closed work is not reopened as if it were new.
Multi-Model AI Access River combinatorics, polarized-range counting, and MDF math benefit from a precise model; Socratic questioning and tilt work benefit from a more conversational one; board-texture diagrams and range grids need image generation. Poker Coach runs on Jenova's multi-model access — OpenAI, Anthropic, Google, xAI, DeepSeek, Qwen, and more — so a close river spot and a written study curriculum are not forced through a single-vendor stack such as ChatGPT (OpenAI-only) or Claude (Anthropic-only). Custom model selection is available on paid tiers.
Intelligent Tool Integration for End-to-End Workflows The production path is a pasted hand history or table talk description → street-by-street analysis with leak updates → exportable artifacts. Google Search and YouTube Search cover live schedules, rake changes, and concept videos; Google Scholar covers game-theory foundations; CSV, PDF, DOCX, and image generation produce range charts, study plans, session recaps, and range grids. That chain is a coaching desk, not a disposable chat.
Range Thinking Across Formats, Not Memorized Charts Poker Coach thinks in ranges, not hero hands, and keeps advice tied to 6-max and full-ring cash, MTTs, Sit & Gos, Spin & Gos, Pot-Limit Omaha starting-hand and wrap evaluation, and mixed-game switching discipline. Frameworks such as Janda's bet-sizing theory, Acevedo's Modern Poker Theory, and Tendler's mental-game work are referenced when they apply. Solver numbers for a specific node are not invented.
Online cash-game regulars 6-max and full-ring No-Limit Hold'em grinders who want HUD-informed population reads, table and seat selection, deep-stack planning, and leak work on c-bets, 3-bet pots, and river bluff-catching.
MTT, Sit & Go, and Spin & Go grinders Tournament players who need ICM, bubble dynamics, chip EV versus $EV, pay-jump leverage, short-stack strategy, re-entry and late-reg considerations, and final-table adjustments instead of cash-game defaults on a bubble.
Live cardroom players Live players working on physical tells, table-talk dynamics, live population tendencies, pace, and live-specific bankroll and tipping considerations rather than online HUD jargon.
Pot-Limit Omaha and mixed-game players PLO players focused on starting-hand selection, equity distribution, wrap and draw evaluation, and equity rundown patterns, plus anyone who switches formats and needs to know which rules actually transfer.
Players stuck on tilt or unstructured study Anyone whose volume is tangled with entitlement tilt, results-oriented thinking, or solver time that feels productive but does not fix identified leaks.
Free For exploring Jenova:
Plus ($20/month) For individuals:
Premium ($50/month) For regular users:
Pro ($100/month) For power users:
Max ($200/month) For professionals:
Ultra ($500/month) For heavy professionals:
Enterprise ($1,000/month) For businesses:
All plans are billed monthly with no daily usage caps — your full quota is available immediately on each billing date.
Poker Coach is an AI poker coach that builds independent thinkers rather than players who memorize charts. The agent covers preflop range construction, postflop board-texture reading, GTO concepts such as minimum defense frequency and polarization, exploitative adjustments, ICM, and mental-game work, then scales the math to how you already talk about hands. Sessions can be Socratic drills or direct street-by-street breakdowns depending on whether you need a question or an answer.
Poker Coach keeps a durable player profile, Active Leaks, opponent notes, and drill-performance patterns, so a river leak or a regular's fold-to-raise tendency does not vanish when the thread gets long. Multi-model access — OpenAI, Anthropic, Google, xAI, DeepSeek, Qwen, and more — lets combinatorics-heavy spots, Socratic coaching, and board-texture visuals use different strengths instead of a single-vendor lock-in. Search, CSV range charts, PDF study plans, DOCX recaps, and image generation turn a pasted hand history into a study artifact, not just commentary.
Poker Coach uses Google Search, YouTube Search, and Google Scholar for tournament schedules, platform rake changes, concept videos, and game-theory research, and uses CSV, PDF, TXT, DOCX, and image generation for range charts, bankroll trackers, session templates, study plans, hand-analysis notes, and board-texture or range-grid visuals. The intended flow is input (hand history or spot) → analysis and leak tracking → a file or diagram you can review away from the table. Uploaded hand histories and notes can be treated as ongoing reference content for later sessions.
Poker Coach parses standard exports from PokerStars, GGPoker, Winamax, 888, WPT Global, and similar formats, and also reads informal written recaps. Missing data is requested — positions, stack depths, reads — rather than assumed. Several hands in one batch are scanned for shared leaks before any one board runout is over-weighted.
Poker Coach does not fabricate solver outputs, node-specific percentages, or citations. GTO conclusions are stated as general theory; well-known frameworks can be named when they apply; a number that did not come from a tool result or established theory is not invented. Close spots are labeled as close, and what the decision depends on — villain type, stake, polarization — is specified.
Poker Coach will not guarantee profit, win rate, or a stake move. Variance is treated as inherent, and lines are framed as +EV or −EV given the information you provide, not as promised results. If bankroll rules are being broken, that is addressed directly instead of being ignored in favor of a cooler hand.
Poker Coach starts from a neutral-intermediate baseline and calibrates from terminology and hand-description quality over a few exchanges, without a mandatory "what's your level?" quiz. Beginners who ask whether a hand is "good" without position get pot-odds language and full context requests; players discussing node-locking and unexploitable frequencies get combo counts and MDF. User corrections override any read immediately.
Poker Coach covers No-Limit Hold'em cash (6-max and full ring), MTTs, Sit & Gos, Spin & Gos, Pot-Limit Omaha starting hands, wraps, and equity rundowns, plus fundamental mixed-game adjustments. Tournament work includes ICM, bubble dynamics, chip EV versus $EV, pay-jump leverage, short-stack play, and how third-party stacks change a spot. Cash work includes deep-stack play, table selection, and session management.
Poker Coach treats tilt and burnout as first-order issues, not as a preface to skip. Entitlement, revenge, injustice, winner's, and desperation tilt are named when they show up in your reasoning, along with biases such as recency in opponent modeling and loss aversion in marginal spots. Strategy resumes after the mental-game piece is addressed, not instead of it.
Poker Coach does not provide financial advice beyond format-specific bankroll management and does not diagnose gambling addiction. If play-beyond-means or addiction signs appear, the agent points to professional resources such as the National Council on Problem Gambling or GamCare rather than continuing as if the issue were a preflop leak. Earning-potential promises are out of scope.
Poker Coach runs with full feature parity on web, iOS, and Android, so the same coaching thread, memory, and tools are available away from a desktop. Follow the cardroom's device and phone-use rules; the apps do not change house policy. iOS: https://apps.apple.com/app/jenova/id6747692301 — Android: https://play.google.com/store/apps/details?id=ai.jenova.app.prod
Poker Coach cannot schedule recurring reviews, send weekly plans on a timer, monitor a bankroll between sessions, or alert you to tournament registrations. Study plans, bankroll-tracker CSVs, and session templates can be generated on request during a conversation, then used on your own schedule. Background monitoring and webhook-style workflows are not available.
Poker Coach is part of Jenova, the most powerful AI agent platform in the world. Jenova provides unified access to leading AI models from OpenAI, Anthropic, Google, xAI, DeepSeek, Qwen, and more — alongside intelligent tool integrations for end-to-end task completion. The platform features persistent cross-session memory, automated workflow orchestration, and full parity across web, iOS, and Android. Developed by Azeroth Inc., a New York-based technology company, Jenova serves users in over 70 countries.
For developers, the Jenova Agent API provides a fully-managed agent backend with built-in orchestration, multi-model routing, memory, storage, and MCP tool support — replacing the need to stitch together raw LLM APIs, vector databases, and custom tool servers.
Contact: [email protected]