You will get a poker AI training system with solver insights and live game analysis
Top Rated
Top Rated
Project details
I build poker AI systems that combine solver-level precision with real-time usability. My experience ranges from designing a commercial poker trainer for PureGTO to developing a live, AI-powered RTA system.
With this project, you’ll receive a customized poker training application that helps you study optimal play, analyze ranges, and explore exploitative adjustments; not just a toy bot, but a tool built on principled architecture.
What you get:
- Reliable game state parsing (via OCR / hand histories / APIs).
- Integration with leading solvers (Pio, Simple, or custom engines).
- Action recommendations with range + equity explanations.
- Full Java-based, immutable, domain-driven codebase.
- Source code + documentation, ready for long-term use and extension.
Who this is for:
- Poker coaches who want to give their students GTO-based feedback.
- Training platforms seeking solver + AI explanations in one tool.
- Advanced players wanting a structured study companion.
I don’t ship shortcuts or throwaway scripts. You get a maintainable, production-grade system; ready to extend, analyze, and evolve with your poker needs.
With this project, you’ll receive a customized poker training application that helps you study optimal play, analyze ranges, and explore exploitative adjustments; not just a toy bot, but a tool built on principled architecture.
What you get:
- Reliable game state parsing (via OCR / hand histories / APIs).
- Integration with leading solvers (Pio, Simple, or custom engines).
- Action recommendations with range + equity explanations.
- Full Java-based, immutable, domain-driven codebase.
- Source code + documentation, ready for long-term use and extension.
Who this is for:
- Poker coaches who want to give their students GTO-based feedback.
- Training platforms seeking solver + AI explanations in one tool.
- Advanced players wanting a structured study companion.
I don’t ship shortcuts or throwaway scripts. You get a maintainable, production-grade system; ready to extend, analyze, and evolve with your poker needs.
AI Development Type
Deep Learning, Knowledge Representation, Model Tuning, Recommendation System, Software MaintenanceAI Tools
Deeplearning4j, Keras, NVIDIA AI Platform, OpenCV, PyTorch, TensorFlowAI Development Language
JavaWhat's included
| Service Tiers |
Starter
$450
|
Standard
$1,350
|
Advanced
$4,500
|
|---|---|---|---|
| Delivery Time | 7 days | 14 days | 30 days |
Number of Revisions | 1 | 2 | 3 |
AI Model Integration | |||
Detailed Code Comments | - | ||
Knowledge Graph | - | - | |
Model Documentation | - | ||
Ontology | - | - | |
Source Code | |||
Taxonomy | - | - |
Optional add-ons
You can add these on the next page.
Additional Revision
+$90
Extra Solver Integration
(+ 2 Days)
+$450
66 reviews
(64)
(1)
(0)
(1)
(0)
This project doesn't have any reviews.
GC
Gta C.
May 30, 2026
I want to build a desktop software that is able to check financial related documents
AP
Alois P.
Apr 20, 2026
Custom Decision Helper – Desktop App
DR
Divya R.
Nov 6, 2025
Participating in an academic survey to reflect on your freelancing experience - P4.1
LS
Lars S.
Jul 18, 2024
Participants for Programming Study with Software Developers on Java API Usability
Very good communication, delivered on time. It was nice working with him!
HA
Husnain A.
Jul 8, 2024
Java inventory Project
About Hiram
Java chaos tamed for good
92%
Job Success
Nairobi, Kenya - 10:21 pm local time
Some things I've worked on:
A JavaFX scheduling app kept putting things in the wrong state whenever users moved through it quickly. Nobody could reliably reproduce it. I traced where the event flow had grown dependent on interaction timing and corrected the sequencing without touching the rest of the system.
A desktop tool comparing invoices against purchase orders was giving different answers for the same documents. The processing steps weren't running in a fixed order, so the result kept changing. Fixing the pipeline made the output consistent.
PureGTO, a poker analytics platform with a live user base, had errors that only appeared under load. I worked on it for eight months, identified the state transitions causing divergence at volume, and stabilized the runtime behavior.
How I work:
Every job starts with a short review, two to three days. You send me the code, a way to run the program, and a description of what's going wrong. I send back a write-up: what's happening, where it's coming from, and a plan to fix it. If I don't see a clear path forward, I say so and stop there.
After that, I work on a copy of your system. Every change gets tested repeatedly before it touches your live version. I add tests around the fix so the same problem doesn't quietly come back later.
Fixing one problem area usually takes one to three weeks after the review. I work on one system at a time and stay until that part is stable.
I've also built systems from scratch: embedded JavaFX software for an industrial client, a financial document processor, a custom decision tool for a desktop application.
“This was arguably my best hiring experience on Upwork. Hiram does his job really well and in time.” ~ Anton Konkevych, Profiwash, after building embedded JavaFX software for industrial equipment.
If your Java program is behaving strangely and you're not sure why, describe what you're seeing in a message. I'll tell you quickly whether it's something I can help with.
Steps for completing your project
After purchasing the project, send requirements so Hiram can start the project.
Delivery time starts when Hiram receives requirements from you.
Hiram works on your project following the steps below.
Revisions may occur after the delivery date.
Requirements Alignment
Gather variant, solver, and platform details. Confirm use cases (training vs. real-time assistance).
Parsing & Game State Extraction
Implement reliable table state recognition (cards, stacks, actions) via OCR/OpenCV or hand history ingestion.