CedarDB Unifies Transactional and Analytical Databases Amid AI Agent Rise

Image: TechCrunch · Source

CedarDB, a Technical University of Munich spinout, offers a database engine supporting both transactional and analytical queries with high efficiency, gaining relevance as AI agents increase demand for integrated data access.

Traditionally, databases have been specialized for either analytical or operational purposes: analytical databases enable businesses to identify trends and generate insights from collected data, while operational databases manage daily transactional records like orders and payments. CedarDB, a database engine developed by a Munich-based startup spun out from the Technical University of Munich, challenges this dichotomy by supporting both types of queries in one system. This approach has become increasingly relevant with the emergence of agentic artificial intelligence, which demands more seamless and rapid access to diverse data types.

CedarDB’s technology originated from the Umbra research project, which was designed to outperform legacy database engines by maximizing modern hardware capabilities for speed and efficiency. The founding team, including five researchers from Umbra, prioritized maintaining compatibility with PostgreSQL—a widely used open-source database system—to reduce switching costs for users and avoid requiring complete app rewrites. This compatibility also aligns well with the logic familiar to large language models (LLMs), which have been trained extensively on PostgreSQL-like structures.

The company initially launched CedarDB Enterprise as an on-premise solution alongside a free Community Edition in May 2025, supported by a €5.3 million seed funding round. While the company hesitated to offer a cloud solution at first, following beta tests with interested users on its waitlist, it recently introduced limited cloud access at TechCrunch Disrupt, aiming to broaden its customer base beyond large enterprises.

This cloud-hosted version positions CedarDB to compete more directly with established data platforms like Databricks, which recently announced LTAP—a similar effort to unify transactions, analytics, streaming, and operational data in a single storage framework. CedarDB’s CEO Moritz Sichert interprets these developments as validation of the market need for integrated database solutions that serve the evolving demands of AI-driven applications.

Currently, CedarDB operates with a 14-person team, focusing on serving the initial batch of cloud users while planning to scale up with additional funding rounds as user adoption grows. The company’s evolution reflects a broader trend in the database industry towards offering unified, efficient data platforms to support the next generation of AI and enterprise workloads.