AI-Powered Chess Analysis Debuts at Bologna Tournament

• Intel-backed agentic AI offers post-game insights and real-time commentary, enhancing strategic understanding for players and viewers alike.

AI Joins the Chessboard in Bologna

At the Alma Mater University Chess Tournament held September 12–14 in Bologna, Italy, a new form of artificial intelligence made its debut alongside traditional chess competition. Teams from 18 universities entered the AI Analysis Room after each match, where a chess-playing agentic AI provided strategic feedback and move optimization. This post-game reflection aimed to help players refine their tactics for future rounds. Spectators also engaged with the AI, gaining access to expert-level commentary that demystified complex plays.

The initiative builds on the idea that chess mastery involves constant improvement, echoing Emanuel Lasker’s famous advice to always seek a better move. By integrating AI into the tournament experience, organizers hoped to make high-level analysis more accessible. The agentic AI served not only as a tool for learning but also as a bridge between players and audiences. Its conversational interface allowed users to explore game dynamics in a more intuitive way.

ShashGuru Blends Language and Strategy

Central to the AI Analysis Room was ShashGuru, an open-source research platform developed by computer science student Alessandro Libralesso. Based on ShashChess—a customized engine derived from the widely respected Stockfish—ShashGuru adds a layer of natural language interaction. It combines Meta’s Llama-3.1-8B language model with ShashChess’s analytical capabilities, running on Intel processors for enhanced performance. Professor Paolo Ciancarini of the University of Bologna described the system as a new approach to human-computer interaction in chess.

The AI operates locally using Intel Core Ultra 200V series processors and OpenVINO™ software, delivering real-time feedback without relying on cloud infrastructure. Between matches, players receive interpretable insights into their own and their opponents’ strategies. For viewers, ShashGuru runs on Intel Xeon 6 processors, optimized for responsiveness and scalability. This setup ensures that both individual users and large audiences can benefit from the AI’s capabilities.

Performance and Learning Beyond the Cloud

Intel’s involvement in the project highlights a shift toward edge computing in AI applications. According to Alessandro Palla, a senior deep learning engineer at Intel, the system is designed to deliver fast, reliable analysis while minimizing errors. Reinforcement learning techniques help reduce hallucinations, making the AI’s suggestions more trustworthy. The goal is to provide a strategic companion that supports both competitive play and broader understanding.

By operating outside the cloud, the AI maintains privacy and reduces latency, which is particularly valuable in live tournament settings. The combination of chess engine precision and conversational AI creates a unique educational tool. Players can revisit their games with detailed feedback, while fans gain deeper appreciation for the sport’s nuances. This approach may influence future tournaments and training environments across the chess world.