Web3 Chatbot
AI agent for on-chain data

Conceptual cover artwork, not an application screenshot.
Overview
A conversational AI assistant that can answer questions about on-chain data and Web3 concepts in plain language — wallet balances, transaction history and protocol basics, without the user needing to read a block explorer.
Features
- Natural-language queries translated into on-chain data lookups
- Supports basic wallet and transaction lookups via a connected provider
- Explains Web3 concepts in plain language for less technical users
- Built with tool-calling so the model fetches real data rather than guessing
Technology
- TypeScript
- Node.js
- Ethers.js
- OpenAI APIs
- Tool calling
- AI agents
- Prompt engineering
- Ethereum
The problem
Block explorers and wallet dashboards are built for people who already understand Web3's mental model. For everyone else, "what's actually in my wallet and what just happened to it" is a harder question to answer than it should be.
Approach
Instead of answering from a static knowledge base, the chatbot is wired with tools that fetch real on-chain data — balances, transactions — and uses the language model purely to understand the question and explain the result, keeping the facts grounded in actual chain data rather than the model's own assumptions.
Outcome
An assistant that treats a wallet or a transaction as something you can simply ask about, aimed at lowering the barrier between "technically on-chain" and "actually understandable."