Cognee, a Berlin-based AI infrastructure startup, has secured €7.5 million in funding to further develop its structured memory layer for AI systems and agents.
The round was led by Pebblebed, with 42CAP participating as a follow-on investor. Additional support came from a mix of existing and new angel investors.
Addressing AI’s Memory Gap
Founded in 2024, Cognee focuses on solving a core limitation of many current AI systems: the lack of persistent, structured memory.
Most AI applications operate in a stateless manner, relying on limited context windows or file-based retrieval methods. While effective for short interactions, these approaches often struggle in complex, long-running, and business-critical workflows where sustained contextual understanding is required.
Cognee’s platform transforms unstructured data into a structured memory layer built on knowledge graphs and semantic models. This enables AI systems to retain context over time, connect related information, and operate with improved reliability in production environments.
Advertisement
From Open Source To Enterprise Infrastructure
Originally launched as an open-source initiative, the company has gained adoption among developers building AI-native applications that require contextual continuity and dependable memory capabilities.
The company now serves a growing base of engineers and enterprises across Europe and the United States, positioning its technology as foundational infrastructure for next-generation AI systems.
Founder and CEO Vasilije Markovic stated that AI systems often fall short not due to insufficient capability, but because they lack the ability to maintain and structure memory over time. Cognee’s objective is to provide that underlying memory layer, enabling AI to move beyond simple text retrieval toward contextual understanding.
About Cognee
Founded in 2024 and headquartered in Berlin, Cognee develops a knowledge and memory engine for AI applications and agents. Its technology converts unstructured data into structured, persistent memory, allowing AI systems to retain context, reason across connected information, and deliver more reliable performance in enterprise environments.
