Signup for Free
Featured

Announcing: Zep Community Edition

Zep Community Edition, open-sourced today, is the first Zep product powered by a Knowledge Graph, allowing you to build more accurate and personalized Agents.

Daniel Chalef

• 2 min read
Announcing: Zep Community Edition
💡
Visit Zep on GitHub: github.com/getzep/zep

Hello Zep Community -

We’ve just open-sourced Zep Community Edition (CE), a memory layer for AI agents that continuously learns facts from user interactions and changing business data. Zep ensures that your Agent has the knowledge needed to accomplish tasks successfully.

A few weeks ago, we shared with you Graphiti, our library for building temporal Knowledge Graphs. Zep Community Edition runs Graphiti under the hood, progressively building and updating a temporal graph from chat interactions, tool use, and business data in JSON or unstructured text.

Zep CE's graph memory offers significantly greater accuracy and flexibility than previous versions of Zep. The graph memory engine will be coming to Zep Cloud in the coming weeks, along with additional enhancements to the service.

Why use a graph?

We believe temporal graphs are the most expressive and dense structure for modeling an agent’s dynamic world, such as changing user preferences, traits, and business data.

Zep reconciles changes in facts as the agent’s environment changes. We use temporal metadata on graph edges to track fact validity, allowing your agent to reason with these state changes:

Fact: “Kendra loves Adidas shoes” (valid_at: 2024-08-10)

User: “I’m so angry! My favorite Adidas shoes fell apart! Puma’s are my new favorite shoes!” (2024-09-25)

Facts:

  • “Kendra loves Adidas shoes.” (valid_at: 2024-08-10, invalid_at: 2024-09-25)
  • “Kendra’s Adidas shoes fell apart.” (valid_at: 2024-09-25)
  • “Kendra prefers Puma.” (valid_at: 2024-09-25)

Several popular memory layers offer agent-powered retrieval and complex multi-level architectures. We've found these to be slow, non-deterministic, and difficult to reason with. Zep’s approach, which asynchronously precomputes the graph and related facts, supports very low-latency, deterministic retrieval.

In addition to an opinionated high-level memory persistence and retrieval API, developers can build custom queries into a user graph. Additionally, a Create, Read, Update, and Delete API offers fine-grained control over data in Zep, and supports compliance with privacy and data regulation.

For Zep Open Source Users

Please see important information in the GitHub README related to ongoing access to Zep Open Source v0.x. We are currently exploring the feasibility of an upgrade path to Zep CE.

-Daniel and the Zep team

Next steps