Meko Documentation

Meko is an agent-native unified data layer for multi-agent systems, where agents work and learn together from shared knowlege and collective memory with full auditability.

Meko offers:

  • Shared knowledge: build shared knowledge over time from conversations, real-time data sources, and slower-changing knowledge bases.

  • Collective memory: learning that compounds across the agents working together, as opposed to just per-agent memory.

  • Auditability of the learning process: connect execution traces to how agents working together process data and knowledge, learn from it, and share that learning.

Meko integrates easily with any agentic framework through MCP servers. It is serverless, multi-tenant (multi-agentic), and optimized for tiering to object stores. It is built on top of a unified distributed PostgreSQL data layer that supports vector, SQL, NoSQL, graph, vector search, and graph search.

   

Quick Start

Sign up for free, connect, and start using Meko with your agents in less than 60 seconds

Get started →
Quick Start

Learn more about Meko

How Meko fits

Where Meko sits in your architecture and how it deploys.

Learn more →

Core concepts

Understand the foundational building blocks of Meko.

Learn more →