AIGame: Architecture of the Platform for Studying AI-Agent Interaction and Immersion

AIGame: Architecture of the Platform for Studying AI-Agent Interaction and Immersion

Egorova E.P. (NRU HSE, Moscow, Russia)

Abstract

This paper addresses the problem of designing a specialized software platform for empirical research in the field of human-AI interaction. The primary objective is to study the interrelationship between the complexity of the game world model and the level of immersion experienced by the player when interacting with intelligent agents. A modular, multi-tiered microservice architecture is proposed, providing the necessary flexibility for conducting variable experiments. Key architectural characteristics include: isolated execution of agent code in Docker containers, clear separation of the Game Loop and the development/testing cycle, and the use of specialized protocols (gRPC, WebSocket) for different types of interaction. The proposed architecture allows researchers to flexibly modify both agent behavior and game world rules while ensuring deterministic data collection for subsequent analysis.

Keywords

immersion; sandbox; agent.

Edition

Proceedings of the Institute for System Programming, vol. 38, issue 6, part 2, 2026, pp. 227-238

ISSN 2220-6426 (Online), ISSN 2079-8156 (Print).

DOI: 10.15514/ISPRAS-2026-38(6)-34

For citation

Egorova E.P. AIGame: Architecture of the Platform for Studying AI-Agent Interaction and Immersion. Proceedings of the Institute for System Programming, vol. 38, issue 6, part 2, 2026, pp. 227-238 DOI: 10.15514/ISPRAS-2026-38(6)-34.

Full text of the paper in pdf Back to the contents of the volume