This application is built to read poker tables and process the game state, then work in conjunction with a BBN and/or LLM to make profitable decisions, creating an agentic poker player.
I built this to explore automated decision-making systems and demonstrate large-scale software architecture. The system combines computer vision, multiple OCR methods, and AI reasoning to replicate a profesional poker player.
Working prototype of table maker uploaded. Bot prototype is ready, uploading soon. Multiple planned updates to simplify the current codebase once core functionality is proven. A lot of the code works but is still a prototype and has many places for improvement and simplification. I have a list of planned additions, modifications, simplifications I wish to add at the bottom.
There are two separate applications:
- Table Maker (MorningCoffee) - The region mapping and OCR testing tool
- Poker Bot (Coffee) - The actual decision-making system
Originally had coffee-themed names that I'm in the process of changing.
Table Maker allows you to:
- Create JSON table maps by defining regions on poker screenshots
- Explore image processing settings and OCR parameters
- Save and capture images in real time for building and refining models
- Test multiple OCR methods (Apple Vision, Tesseract, CoreML) with confidence tracking
- Export training data for custom model development
Poker Bot handles:
- Reading poker table state through the mapped regions
- Processing game information into structured data for AI consumption
- Making profitable decisions using a pre-trained LLM or Bayesian Belief Network
- Screen Capture Engine - Real-time monitoring of poker interfaces
- Multi-OCR Pipeline - Apple Vision, Tesseract, and custom CoreML models
- Region Mapping System - Visual interface for defining table elements
- Confidence Tracking - Reliability scoring across all recognition methods
- Training Data Collection - Automated dataset generation for model improvement
- Game State Parser - Converts visual data into structured game information
- LLM Integration - Uses pre-trained language models for strategic reasoning
- Bayesian Networks -
- Game Types - NLH (8-MAX), More to Come later
- Multi-modal OCR with automatic fallback systems
- Custom CoreML models for card recognition (A, K, Q, J, T, 9-2)
- Real-time confidence scoring and accuracy tracking
- Preprocessing optimization for challenging poker interface elements
- LLM reasoning for complex strategic situations
- BBN probability analysis planned, for uncertainty management
- Profitable play optimization focused on long-term value generation
- Rapid Collection Mode - Batch capture with automatic tagging
- Training Data Export - TIFF images with ground truth for OCR training
- CoreML Dataset Generation - Raw images organized for custom model training
- These are 3 seperate pieces that I will combine later into a single simplified interface
- macOS (silicon)
- Xcode for compilation
- Screen Recording & Accessibility permissions
- Homebrew package manager
brew install tesseractI will be updating this potential problems you could run into, and howto fix them