Decentralized AI (DAI)
DAI is a decentralized AI platform built in Rust, designed to run AI tasks over a peer-to-peer network with blockchain-backed data integrity. It enables secure, distributed computation and collaborative AI processing across multiple nodes — with no central server in the middle.
Key contributions:
- Designed and implemented a peer-to-peer protocol in Rust for secure node-to-node communication and distributed AI task coordination.
- Developed a blockchain-based system for recording task submissions, results, and consensus verification across nodes.
- Implemented message handling for blocks, transactions, peers, and pings, ensuring real-time network updates and synchronization.
- Integrated asynchronous networking and connection pooling to manage multiple node connections simultaneously using Tokio.
The core idea: when compute is spread across untrusted peers, the blockchain becomes the shared source of truth — who submitted what, which results were verified, and what the network agreed on. It’s the same problem consensus systems have always solved, applied to AI workloads.