AgentRecall
Persistent, intelligent memory for AI agents across sessions
AgentRecall is an open-source SDK that gives AI agents persistent memory using a Neo4j-powered graph database, vector semantic search, and AI-powered entity extraction. It supports both self-hosted and cloud deployment modes, allowing developers to store, search, and traverse memories across conversations. The SDK integrates with tools like Claude Desktop via MCP and supports multi-agent architectures with isolated namespaces.
Developers install the SDK, store memory entries after each interaction, and retrieve relevant context via semantic search or graph traversal — backed by Neo4j and vector embeddings.
AI developers building agents or multi-agent systems
Background.
- Status
- launched
- Business model
- freemium
Similar projects.
Editorial take on the space this project sits in — momentum signals, adjacent moves, our call on whether the wedge is real. Get pinged when we publish a new read or when the landscape shifts.
Have a take on this space?
Tell us what you’d build differently, where you think the incumbents miss, or what we’ve gotten wrong about this project. Comments + reactions are coming soon.