Skip to main content

Welcome to ChessAIthon

Introduction​

Welcome to the technical documentation for the ChessAIThon project, an initiative designed to bridge the gap between vocational education and the evolving demands of the digital workforce. This technical memory specifically details the software development phase, outlining the methodologies, tools, and processes employed to bring the project's innovative vision to life.

The ChessAIThon project, deeply rooted in the belief that strategic gaming like chess can cultivate essential skills such as critical thinking, pattern recognition, and decision-making, aims to integrate the practical application of Artificial Intelligence (AI) into Vocational Education and Training (VET) curricula. Our consortium recognized the critical need for VET institutions to adapt and equip students with robust coding, data management, and AI competencies, areas often overlooked in traditional IT curricula.

This project directly addresses the VET sector's priority of "Adapting vocational education and training to labour market needs" by providing a balanced mix of vocational skills through strategic gaming and understanding AI implementation. Furthermore, it contributes to "Innovation in vocational education and training" by engaging students in a dynamic learning environment that applies AI to chess through an interactive interface. The project also aligns with the "Addressing digital transformation through development of digital readiness, resilience and capacity" horizontal priority, stimulating the development of digital pedagogical skills by purposefully incorporating digital technologies, AI, and chess.

The core of ChessAIThon lies in its unconventional approach: integrating chess logic into the curriculum to develop foundational skills necessary for using AI. The logical and algorithmic thinking inherent in chess seamlessly aligns with the cognitive processes integral to coding and data analysis, making it an ideal complement to traditional programming education.

This document will detail the development of key components, including:

  • A comprehensive online database: This dynamic repository will feature real-life chess scenarios and allow users to propose solutions, ultimately aggregating student and player contributions to train an AI. The system will displaying boards, legal moves and enable data export for archiving and version control.

  • The Chess Artificial Intelligence Hackathon platform: This innovative competition will challenge students to use automated tools or create their own to train an AI with their documented chess moves, fostering a holistic understanding of chess strategy, computational thinking, data structures, version control, AI models, and cloud computing.

By focusing on these practical applications, ChessAIThon aims to provide VET students with a more comprehensive learning experience that goes beyond basic programming, incorporating essential aspects such as effective data management, version control, and a deeper exploration of AI concepts not commonly encountered at this educational level. This technical memory serves as a testament to the project's commitment to preparing students for the technical demands of contemporary workplaces and fostering a new generation of skilled professionals in AI and frontend development.

Objectives​

Curriculum Integration & Teacher Support: Provide tools to integrate chess-based learning for developing critical thinking, problem-solving, and creativity. This includes functionalities for teaching coding principles through chess scenarios, fundamental AI and Machine Learning concepts, data structures, and the utilization of public datasets like Kaggle for AI training. The software will also facilitate teaching the use of Large Language Models (LLMs) for chess and coding problems, and version control for chess scenarios.

Interactive Learning Platform: Develop an online platform with a visually appealing chessboard interface. This platform will enable users to propose and engage with real-life chess scenarios. It will utilize the Chess.js library for displaying all legal moves, validating user moves, and storing valid moves in a database.

Data Management & Analysis: Implement robust data handling capabilities, including the representation of chess with various computer file formats like Portable Game Notation (PGN), JSON, CSV, Forsyth-Edwards Notation (FEN), Universal Chess Interface (UCI), and Standard Algebraic Notation (SAN). The system will support exporting scenarios and moves to CSV or similar formats for archiving and version control. It will also facilitate data analysis principles based on chess data.

AI Training & Competition Framework: Provide methodologies and software for students to work with solution datasets to train Artificial Intelligence models. This includes enabling students to fine-tune AI training and measure performance. The software will culminate in a chess and AI competition where student-trained AIs compete, reflecting the collective intelligence of each group.

Version Control Integration: Utilize version control tools such as Git, specifically through platforms like GitHub, to ensure data continuity and seamless collaboration among partners for storing and sharing datasets of chess problem-solving challenges and solutions.

ChessAIthon is the name of the complete project. As a pun, we named ChessNet the NN base of our project. ChessMarro is the name of the AI (ChessNet+Deploy). ChessMinds is the name of the web frontend to play with ChessMarro and to create datasets.

Open Sourcing the project​

This strategy represents a high-standard framework for Open Science and Technical-Vocational Education and Training (VET). By distributing assets across specialized platforms, the project ensures high availability, technical transparency, and a frictionless learning curve.

1. Models amd Datasets: Hugging Face & Kaggle​

The separation of model weights from the source code is a fundamental practice in modern AI. Hosting on Hugging Face provides a robust API for programmatic access, while Kaggle offers a direct interface for data science experimentation.

  • Academic Justification: This dual-hosting strategy prevents "Data Siloing." It allows for independent verification of the neural network’s architecture and weights, facilitating peer review and "Transfer Learning" research.
  • VET Educational Benefit: Students gain experience with Model Registry concepts. They learn to manage large-scale binary files separately from source code, a critical skill in MLOps (Machine Learning Operations).

2. Code & Documentation: GitHub & GitHub Pages​

The project utilizes GitHub for version control and Docusaurus (via GitHub Pages) for technical documentation. Docusaurus transforms Markdown—a lightweight markup language—into a structured, searchable, and web-optimized manual.

  • Academic Justification: Documentation-as-Code ensures that the "Technical Manual" evolves at the same rate as the software. By making the documentation human-readable and machine-parsable, we ensure long-term project sustainability.
  • VET Educational Benefit: It exposes students to CI/CD (Continuous Integration/Continuous Deployment). They see how a commit in a repository triggers an automated build that updates the public documentation, mirroring professional software engineering workflows.

3. Notebooks: Kaggle & GitHub​

By providing "Ready-to-Run" Jupyter Notebooks on Kaggle, the project removes the hardware barrier.

  • Academic Justification: This ensures Total Reproducibility. In the "Replication Crisis" of modern AI, providing the exact environment and compute used for training is the only way to validate scientific claims.
  • VET Educational Benefit: Students often lack high-end hardware (GPUs/TPUs). Kaggle’s free compute allows students from all socio-economic backgrounds to execute complex MCTS simulations and neural training, promoting equity in technical education.

4. Deployment: Hugging Face Spaces (MCTS)​

curl -X POST "https://jocasal-chessAIthon.hf.space/predict"      -H "Content-Type: application/json"      -d '{"fen": "rnb1k1nr/pppp1ppp/5q2/2bP4/8/5N2/PPP1PPPP/RNBQKB1R w KQkq - 3 6", "simulations": 4000}'

{"move":"b1c3","visits":1956,"alternatives":[["e2e3",945],["e2e4",583],["c1g5",115]]}

Deploying the MCTS engine via Hugging Face Spaces (using FastAPI) creates a functional bridge between the backend logic and the end-user.

  • Academic Justification: This serves as a "Black-Box Testing" environment. It allows researchers to interactively probe the model for biases or strategic weaknesses in real-time without local configuration.
  • VET Educational Benefit: It teaches Web-Service Integration. Students learn how a Python-based AI backend can be exposed through a web interface, a core competency for modern Full-Stack AI developers.

5. Web Application: Vercel (ChessMinds Webpage)​

The use of Vercel for the project's application to play chess provides a professional, high-performance entry point.

  • Academic Justification: Separation of Concerns. By decoupling the "Project Application" (Vercel) from the "Technical Documentation" (GitHub Pages), the project maintains a clear hierarchy of information for different stakeholders.
  • VET Educational Benefit: Students observe the importance of Frontend Performance and CDN (Content Delivery Network) distribution. They learn that user experience (UX) is as vital as the underlying algorithm for project adoption.

The ChessAIthon distribution model is not merely a set of links; it is a coordinated Technical Ecosystem. It adheres to the FAIR Principles (Findable, Accessible, Interoperable, and Reusable). For VET institutions, this project provides a holistic "Blueprint" for how modern AI projects are built, documented, and deployed in the industry today.