Yugabyte launches Meko platform to give AI agents shared memory and traceability

Yugabyte announced the launch of Meko, a data‑infrastructure platform designed to provide persistent memory, shared knowledge and traceability for multi‑agent AI systems. CEO Karthik Ranganathan said, “Your agentic systems are only as good as the state and the data that you feed them,” and noted that current agents lack the ability to share context, reasoning and prior knowledge, leading to repeated work and higher token consumption.

Bitweave is an open‑source hybrid retrieval engine written in Rust that achieves sub‑millisecond query latency while keeping the active RSS memory footprint under 1.1 MB. It uses zero‑copy memory‑mapped files and SIMD‑accelerated 1‑bit Hamming‑distance filtering to handle high‑dimensional (e.g., 1536‑dim) vectors, making it suitable for local retrieval‑augmented generation, edge agents and serverless AI pipelines.