TORY

TORY is a multi-agent AI system that provides comprehensive tokenomics analysis for Web3 projects. It uses Fetch.ai's uAgents framework to orchestrate specialized agents that gather on-chain data, analyze token unlock schedules, evaluate financial metrics, and generate actionable investment insights.

Crypto investors need to evaluate tokenomics across hundreds of projects, but the data is scattered across blockchain explorers, project docs, and social channels. Manual analysis is time-consuming and error-prone. The challenge was building an autonomous system that could gather, normalize, and analyze this data reliably using multiple AI agents working in coordination.

I designed a multi-agent architecture using Fetch.ai's uAgents framework, where specialized agents handle different analysis domains: one agent scrapes tokenomics data, another tracks unlock events, and a third generates AI insights using asi1.ai. The agents communicate via the Agentverse protocol. The frontend is built with Next.js on Vercel, and the FastAPI backend on Railway coordinates agent requests and caches results.

Built a functional multi-agent system that delivers comprehensive tokenomics reports in under 30 seconds. The distributed agent architecture allows parallel data gathering, making it significantly faster than sequential API calls. Recognized at the Fetch.ai hackathon for its novel use of the uAgents framework.