AI app builders who need persistent, personalized memory across user sessions and agents.

A fast, scalable memory engine and API that stores and retrieves user context for AI applications, deployable as a hosted service or self hosted.

Accessed via a cloud API or run locally as a self hosted open source deployment.

Use when an agent or chat product needs to remember facts across conversations without building a custom retrieval pipeline.

It gives applications long term, queryable memory that improves personalization without managing a vector database directly.

Built on vector search and embedding infrastructure, with SDKs for common LLM frameworks.

Open source core, MIT licensed, with a managed hosted tier that requires an API key.