Multi-Agent Reinforcement Learning Simulator

A simulation environment where multiple agents learn via Q-learning, trainable and viewable from the browser.

What it does

• Designed the Q-learning training pipeline, agent decision logic, and simulation environment.
• Architected a modular FastAPI backend exposing simulation state through a REST API.
• Built a training state machine managing agent lifecycle and Q-table persistence.
• Developed a React (Vite) frontend to run and visualize agent behavior in real time.

Tags