This is reimaging what it means to have project specific blockchain analytics. This is a completely decentralized analytics platform. When a user logs in, they specify which contract addresses relate to a given DApp application. From there the system dynamically crawls and indexes all relevant information for later retrieval. The stored data makes it easier to do cohort analysis, user retention testing and app to app user base correlation tests. At the heart is a pioneering decentralized MongoDB inspired no-SQL data storage system. Anyone who has ever deployed a smart contract knows how annoying it can be to try and find app specific user data. That is solved now! For more information see:

MeshLink showcase

How it's made

The system is broken up into three parts (1) the crawler (2) the backend api and (3) the decentralized front-end. Within the crawler it pulls in blockchain data using Covalent for a large amount of transaction data for a given address and on top of that, it uses RPC nodes provided by Pokt. The system's database can be thought of as a decentralized MongoDB. During the data ingestion process the system does computations on all the transaction and address data and then breaks that up into separate JSON files split across the Storj network. This allows the front-end to call IPFS nodes instead of pinging the APIs to recompute everything. On the front-end users sign in with WalletConnect and thus have their apps saved to their address.