Berlin-Based AI Infrastructure Startup Cognee Raises €7.5 Million To Expand Its Enterprise Memory Platform

The round was led by Pebblebed, with 42CAP joining as a follow-on investor. A group of existing and new angel investors also took part in the funding.

Founded in 2024, Cognee develops core AI infrastructure for developers and enterprises worldwide. The company addresses a key limitation of current AI systems: the absence of long-term, structured memory.

Most AI applications rely on stateless methods such as file-based retrieval or short context windows, which can struggle with complex, ongoing, and business-critical workflows.

Read Also - Tingit Secures €1.5M To Expand AI-Powered Repair Platform Across Europe

Cognee converts unstructured data into a persistent, structured memory layer built on knowledge graphs and semantic models. This approach allows AI systems and agents to retain context over time, reason across interconnected data and reduce hallucinations in production environments.

Originally launched as an open-source project, Cognee has gained strong traction among the global developer community. Thousands of engineers use the platform to build AI-native applications that require reliable memory and contextual awareness.

Although headquartered in Berlin, Cognee operates across Europe and the United States, focusing on delivering foundational AI infrastructure to a global market.

“AI systems today don’t fail because they aren’t powerful enough,” says Vasilije Markovic, Founder and CEO of Cognee. “They fail because they don’t remember. We’re building the memory layer that allows AI to understand context, not just retrieve text.”

About Cognee

Cognee builds an open-source knowledge and memory engine for AI applications and agents. Its technology transforms unstructured data into structured, persistent memory, enabling AI systems to retain context, reason across information and operate more reliably. Cognee helps developers create AI-native applications with long-term memory and improved performance.mory and improved performance.