ArchiveNet

Where AI memories live forever- A decentralized semantic memory platform powered by blockchain and vector search

ArchiveNET is a revolutionary decentralized memory management platform that combines the power of AI embeddings with blockchain permanence. Built on Arweave, it provides enterprise-grade semantic search capabilities through advanced vector database technology, enabling applications to store, search, and retrieve contextual information with unprecedented permanence and accuracy.

- Semantic Memory: AI-powered contextual memory storage and retrieval
- Blockchain Persistence: Permanent storage on Arweave blockchain
- Vector Search: State-of-the-art HNSW algorithm for similarity search
- High Performance: O(log N) search complexity for millions of vectors
- Decentralized: No single point of failure or censorship
- Rich Metadata: Comprehensive metadata support for enhanced search
- Enterprise-Ready: Production-grade API with authentication and monitoring

ArchiveNET is a comprehensive monorepo consisting of four main components:

The world's first decentralized vector engine built on Arweave blockchain, implementing the Hierarchical Navigable Small Worlds (HNSW) algorithm for approximate nearest neighbor search.

- HNSW algorithm with O(log N) complexity
- Blockchain-based persistence via HollowDB
- Protobuf encoding for efficient storage
- Database-agnostic interface
- Handles millions of high-dimensional vectors

A robust Express.js API service providing semantic memory management with AI-powered search capabilities.

- Express.js with TypeScript
- Neon PostgreSQL with Drizzle ORM
- EizenDB for vector operations
- Redis for caching
- JWT authentication
- Comprehensive validation with Zod

A modern Next.js application providing an intuitive interface for memory management and search operations.

- React-based UI with TypeScript
- Real-time search capabilities
- Memory visualization
- User dashboard
- Responsive design

The central Model Context Protocol (MCP) server that orchestrates memory operations and provides intelligent context management.

- Node.js 18+
- Docker & Docker Compose
- PostgreSQL database
- Redis server
- Arweave wallet

git clone https://github.com/s9swata/archivenet.git cd ArchiveNET

# API Setup cd API npm install npm run build npx drizzle-kit push npm run dev # Frontend Setup cd ../client npm install npm run dev

Create.envfiles in respective directories:

DATABASE_URL=your_postgres_url REDIS_URL=redis://localhost:6379 JWT_SECRET=your_jwt_secret ARWEAVE_WALLET_PATH=./data/wallet.json

// Store a memory const response = await fetch("/api/memories", { method: "POST", headers: { "Content-Type": "application/json" }, body: JSON.stringify({ content: "Project discussion about AI integration", metadata: { project: "AI-Platform", priority: "high" }, }), }); // Search memories const searchResults = await fetch("/api/memories/search", { method: "POST", headers: { "Content-Type": "application/json" }, body: JSON.stringify({ query: "AI project discussions", limit: 10, }), });

- Fork the repository
- Create a feature branch (git checkout -b feature/amazing-feature)
- Commit your changes (git commit -m 'Add amazing feature')
- Push to the branch (git push origin feature/amazing-feature)
- Open a Pull Request

- Developer Guide- Comprehensive API documentation
- Backend Integration- Backend implementation examples
- Eizen Engine- Vector database engine details
- HNSW Guide- Algorithm implementation details

This project is licensed under the MIT License - see theLICENSEfile for details.

good-wook is a free conversational front end to Phish's show history, usable from Claude. Ask in plain English about setlists, song gaps and history, jam charts, venue stats, and upcoming shows. Every answer is fetched live from the Phish.net API and credited to Phish.net; nothing is scraped or stored.