Centurion – AI-driven HH Exploit Tool (Post-Game, No HUD) - Looking for Support
Centurion – AI-driven HH Exploit Tool (Post-Game, No HUD) - Looking for Support

Centurion – AI-driven HH Exploit Tool (Post-Game, No HUD) - Looking for Support

Centurion – AI-driven HH Exploit Tool

I’ve been developing a side project for several months to plug a specific hole in the poker software stack. Trackers and HUDs churn out endless raw stats, but there’s no simple, reliable tool that converts those numbers into clear, actionable exploits you can actually use at the table or in coaching.

Centurion
is built for three groups: coaches who need ready-to-use leak reports for students, staking stables that want fast villain breakdowns across large HH pools, and serious players who want higher-value post-session review without digging through stat soup.

Right now I’m getting the first genuinely useful results, but there’s still a fair bit of finetuning to do. The core will be an AI model trained specifically for this task so the output is concise, context-aware, and tailored to poker decision-making rather than generic NLP fluff.

In addition to single-villain reports, Centurion will run population-analysis and tendency comparisons. That means it will flag where a player deviates from typical pool behavior, surface directionally important tendencies (e.g., fold-to-steal, 3-bet frequency, c-bet patterns), and translate those differences into concrete exploit recommendations. The roadmap also includes automated session/database analysis (MDA) that you can point at a hand-history dump and get an easy-to-read summary of common leaks, prioritized fixes, and suggested lines based on modern standards and stable workflows.

The project is being built out of Austria, with development and operations based here.

What it does

  • Upload raw HHs (single files or bulk .txt/.hh/.zip)
  • Select a villain or group from the database
  • Stats are scanned against solver-derived priors (e.g., PioSolver/GTO+ benchmarks) and population tendencies to highlight position/street-specific leaks
  • Output = plain-English exploit cards that always include population comparisons, e.g. “Folds BB vs steal 82% (pool avg ~58%) — open wider SB/BTN” or “Flop c-bet 78% vs pool 68% / Turn 22% vs pool 50% — float wider, attack turn”
  • Generate PDF reports with filters (position, stack size, date range) and optional CSV export
  • Run automated database analysis (MDA) for prioritized leaks, recurring patterns, and adjustment suggestions
  • Works across NLH, PLO, cash games, and MTTs
  • Native PC app with local-first processing (not browser-based). Optional secure server processing is available only for very large batches or team workspaces.
  • Never a live HUD or in-game overlay — strictly a post-game review and coaching tool


Why bother

Most trackers drown you in raw stats and noise, leaving players guessing how to actually use the numbers. Centurion flips that around: it highlights leaks, compares them to population tendencies, and translates the data into simple, actionable insights that anyone can understand and apply immediately.

State of project

  • Desktop alpha MVP (native PC app) available; lightweight web demo for quick trials.
  • Local parsing works on demo HHs; early testing with a small staking Discord produced positive feedback (“finally usable leaks instead of endless stats”).
  • Next steps: productionize the multi-site parser (PokerStars/ACR/GG/888/Party/iPoker), integrate population/solver comparisons into the report output, implement PDF/CSV export and team workspaces, add optional secure server processing for large batches, wire subscription billing, and finish UX polish + QA.

Pricing

Free plan: up to 1,000 hands/month.
Pro: €29/month, unlimited analysis.
Team licenses available for stables.

Raise

Seeking funding and strategic support to move Centurion from MVP to full product. Target raise is in the low five-figure range, depending on scope and partner structure. Funds will cover multi-site HH parser development (Stars/ACR/Party/iPoker), desktop app packaging, secure PDF/CSV export, server/infra costs, and initial marketing to coaches and stables.

Backer Packages

  • €500 – Lifetime Pro: one-time support = unlimited hands & features forever.
  • €2,000 – Coach/Stable Pack: Lifetime Pro + 10 transferable Pro accounts.
  • €5,000 – Partner Pack: Lifetime Pro + 25 transferable Pro accounts + input on roadmap.
  • €10,000+ – Revenue Share Tier: All of the above + % of subscription revenue (negotiable, capped + time-limited, e.g. 2% for 24 months).

I’m happy to answer any questions directly in this thread or by PM. Constructive feedback is very welcome, and I’m open to serious discussions with anyone interested in supporting, testing, or backing the project.

29 September 2025 at 12:36 AM
Reply...

1 Reply



any feedback to the idea? 😀 kickstarter campaign coming soon <3

Reply...