Wind Network

Wind Network is a decentralized data indexing and storage infrastructure for Solana that enables the development of advanced AI applications. Our platform processes and indexes blockchain data in real-time through a distributed network of nodes, providing structured data formats optimized for machine learning models and LLMs to train on Solana data.

Key Features

• AI-Ready Data Formats: Structured blockchain data optimized for machine learning models and LLMs
• Decentralized Architecture: Powered by libp2p for robust peer-to-peer communication
• Real-time Data Streams: Process and index Solana blockchain data as it happens
• LLM Training Data: High-quality datasets for training language models on blockchain activity
• Analytics Infrastructure: Foundation for building custom analytics and insight applications
• Enhanced Security: Protected by Jito restaking with economic guarantees for data integrity
• High Performance: Optimized for throughput and low latency
• Horizontal Scaling: Distributed design for unlimited scalability

AI Application Use Cases

• DeFi Analytics Tools: Build applications that analyze market data, monitor liquidity, and track transactions
• NFT Intelligence: Create tools for collection analytics, rarity insights, and trading pattern detection
• Gaming Analytics: Develop applications for player behavior analysis and in-game economy monitoring
• Developer Platforms: Build tools for smart contract usage metrics and network health monitoring
• Research Applications: Enable historical data analysis, trend identification, and market intelligence

Getting Started

Run the development server for the documentation site:

npm run dev
# or
pnpm dev
# or
yarn dev

Then open http://localhost:3000 in your browser.

Project Structure

• crates/windexer-api: Public API interfaces and types
• crates/windexer-common: Shared utilities and common types
• crates/windexer-geyser: Solana Geyser plugin implementation
• crates/windexer-jito-staking: Jito restaking and NCN integration
• crates/windexer-network: libp2p networking and consensus
• crates/windexer-store: Data persistence and caching layer
• crates/windexer-indexer: Data indexing and formatting for AI consumption
• crates/windexer-analytics: Data pipelines for analytics applications

Documentation

Our documentation is built using Fumadocs and provides comprehensive guides on:

• Network architecture
• Data formats for AI
• API reference
• Getting started guides
• Security model

License

This project is licensed under the terms of the license included in the repository.