CedarDB launches cloud database service in 2026, targeting companies that need both transactions and analytics
CedarDB’s cloud database entered limited availability in October 2026 after beta testing with waitlisted users. The Munich startup targets both transactional and analytical workloads, with PostgreSQL compatibility designed to reduce migration costs.
German database startup CedarDB announced limited availability of its cloud-hosted database product in October 2026 at TechCrunch Disrupt, where it is competing in the Startup Battlefield competition alongside 20 other startups.
CedarDB, a spinout of the Technical University of Munich, is built on Umbra, a research project designed to outperform legacy database engines by making better use of modern hardware. Its five co-founders are researchers from that project. The company is headquartered in Munich and currently employs 14 people.
Unlike most databases, which are built for either analytical workloads — spotting trends and answering questions — or operational workloads such as recording transactions, CedarDB is designed to handle both. The company previously launched CedarDB Enterprise, an on-premise product, and CedarDB Community Edition, a free version, in May of the prior year, alongside a €5.3 million (approximately $6 million) seed funding round. The cloud offering was held back while the team conducted beta testing with waitlisted users before announcing limited availability.
CedarDB is also built to be compatible with PostgreSQL, an open source database system. According to co-founder and CEO Moritz Sichert, this keeps switching costs low for customers, since they do not need to rebuild existing applications. Sichert also noted that PostgreSQL compatibility aligns with the logic large language models are most familiar with from their training data.
The rise of agentic AI has added relevance to CedarDB’s approach, Sichert told TechCrunch, as enterprise companies with large stores of transactional data increasingly need to answer analytical questions about that data quickly and in natural language. Sichert pointed to Databricks launching its LTAP data architecture in June 2026 — aimed at unifying transactions, analytics, and operational data — as a signal that demand for this type of unified database is growing.
Sichert said the company plans to serve its first cloud users with its current team, but expects to hire and raise a follow-on funding round once it gains more traction.