Ravenraven
Documentation

Welcome to Raven Protocol

The decentralized memory layer for AI agents. Give your agents the power of permanent, encrypted recall across conversations.

What is Raven?

Raven is a multi-tenant persistent memory layer for LLM/AI agents backed by Walrus blob storage, Sui blockchain, and client-side encryption.

Key Features

  • Multi-tenant architecture - Developers integrate the API; their end-users get isolated memory
  • Conversation-scoped memory - Persistent context across tenant → user → conversation hierarchy
  • AES-256-GCM encryption - All blobs encrypted before storage with per-tenant keys
  • Walrus blob storage - Decentralized, resilient data storage
  • Sui on-chain metadata - Ownership, retention, and verification tracking
  • Agent-side decryption - Autonomous context retrieval without compromising security

How It Works

Raven provides a simple REST API that your AI agents use to store and retrieve conversation memory. Here's the high-level flow:

1

Register as a Tenant

Sign up and receive your API key

2

Create Users

Register your end-users in the system

3

Start Conversations

Create conversation sessions for context isolation

4

Store Memory

Ingest user messages and agent responses

5

Retrieve Context

Query relevant memory for your agent's responses

Quick Example

Here's a quick example of storing and retrieving memory:

Store Memorybash
curl -X POST http://localhost:3000/api/v1/memory/ingest \
  -H "Authorization: Bearer rk_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "user_id": "usr_xxx",
    "conversation_id": "conv_xxx",
    "user_message": "Remember that I prefer TypeScript",
    "agent_response": "Got it! I will use TypeScript for code examples."
  }'
Retrieve Contextbash
curl -X POST http://localhost:3000/api/v1/memory/query \
  -H "Authorization: Bearer rk_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{
    "user_id": "usr_xxx",
    "conversation_id": "conv_xxx",
    "query": "What are the user preferences?"
  }'