OpenBrain FAQs

Question 1

What is OpenBrain?

Accepted Answer

OpenBrain is a self-hosted AI memory layer designed for intelligent agents. It provides shared, persistent memory through typed knowledge tables, advanced hybrid search, and seamless session handoff for continuous context across various AI tools.

Question 2

How does OpenBrain manage different types of AI memories?

Accepted Answer

OpenBrain utilizes distinct typed memory tables for `rules`, `facts`, `incidents`, and `tasks`. Each type has specific lifecycles and management tools, optimizing how AI agents store, retrieve, and utilize information based on its nature.

Question 3

What makes OpenBrain's search capabilities unique?

Accepted Answer

OpenBrain features powerful hybrid search, combining `pgvector` for semantic (cosine similarity) search with Postgres BM25 keyword search. This blend provides highly relevant results, configurable to your specific needs.

Question 4

Can OpenBrain integrate with existing AI agents or clients?

Accepted Answer

Yes, OpenBrain is built as an MCP (Memory Control Protocol) compatible server. It's designed to seamlessly integrate with various AI clients, including Claude Code, Claude.ai, ChatGPT, Cursor, and any other MCP-over-HTTP compatible tool.

Question 5

What are the deployment options for OpenBrain?

Accepted Answer

OpenBrain is a self-hosted solution that can be easily deployed using Docker or integrated into an Onramp setup. It runs on a Postgres database with pgvector, providing flexibility and control over your AI memory infrastructure.

OpenBrain

OpenBrain

Key Features

Use Cases

Key Features

Use Cases