Eli5NearTx

Screenshots

Problem Statement

This tool aims to make blockchain transactions more understandable for users, regardless of their technical background. To do that, we leverage LLMs (in this case OpenAI's gpt-4o) to analyze transaction data. With the help of external data sources like the PikesPeak API, we enrich on-chain data with other information about the addresses involved in the transaction, which gives the LLM a better understanding of the transaction context, and the ability to infer the intentions behind the transaction.

Solution

The project uses a simple HTML + CSS + JS frontend to interact with a python (Flask) backend. The backend uses two APIs:PikesPeak API to fetch transaction information using the transaction hash, and also to fetch general data about all addresses involved in the transaction (affiliations with web3 projects, net-worth, verticals like DeFi/NFT...)OpenAI's "gpt-4o" model to analyze the transaction and account data and provide a good overview of what is happening within the transaction