A decentralized multi-agent system enabling autonomous governance, negotiation, and conflict resolution within a digital economy.
Developed as a rapid prototype in 24 hours during the Fusion Hackathon 2025 to demonstrate self-organizing governance powered by reinforcement learning and adaptive intelligence.
To design and implement a self-evolving governance framework where reinforcement learning (RL) agents autonomously propose, enforce, and evolve governance rules — such as token policies, transaction regulations, and market mechanisms.
The system exhibits emergent economic behaviors, including automated negotiation, consensus formation, and reputation-based trust — all without centralized control.
-
🗳️ Dynamic Governance System
Agents propose and modify rules using Reputation-Weighted Quadratic Voting, ensuring fairness and proportional influence. -
🤝 Autonomous Conflict Resolution
Disputes are resolved via game-theoretic negotiation protocols (Nash Bargaining), fostering stability and cooperation. -
🧬 Adaptive Agents via Reinforcement Learning
Agents continuously learn and adapt to evolving policies and market conditions using PPO in Ray RLlib. -
📊 Real-Time Visualization Dashboard
A live interactive dashboard presents governance metrics, agent reputations, proposals, and simulations.
This diagram illustrates the interactions between RL agents, the decentralized environment, governance modules, and logging/visualization components.
| Layer | Technologies |
|---|---|
| Multi-Agent Simulation | Ray RLlib, PettingZoo |
| Reinforcement Learning | PPO |
| Governance & Communication | Custom Python Modules, gRPC, Redis |
| Data & Visualization | Supabase, PostgreSQL, Dash, Plotly |
| Infrastructure & Automation | GitHub Actions |
- Python 3.10+
# Clone the repository
git clone https://github.com/Harsh-4210/Self_Evolving_Multi_Agent_Governance.git
cd Self_Evolving_Multi_Agent_Governance
# Create and activate virtual environment
python3 -m venv venv
# Linux/macOS
source venv/bin/activate
# Windows PowerShell
.\venv\Scripts\Activate.ps1
# Install dependencies
pip install -r requirements.txt
This project is licensed under the MIT License © 2025 Yash Doke, Viraj Jadhao, Harsh Jain, Bhumi Sirvi. See the LICENSE file for details.