🤖 NFT2VEC 🤖

Problem: Data scientists come to hackathons with big ideas. They spent 2 days trying to fetch the data from the blockchain. In the end they don’t have anything working / to submit.

Solution: Select your data provider (dune, streamingfast, covalent). Collect your data out of the box with a few clicks. Train the Machine Learning Model. Get data insight at your fingertips

1. Used streamingfast from the graph for our first prototype to fetch data and store it locally.
2. Used dune to fetch data from lens, ens
3. Used covalent to fetch 22 millions of NFT transactions

After data collection, we created a data pipeline to convert the transaction data to a graph. Finally we visualise the data with Tensorflow Tensorboard by reducing the dimensions of the embeddings.