CedarDB wants one database to handle both transactions and analytics, as AI agents raise the stakes
CedarDB, a Munich database startup, is building one system for both operational transactions and analytical queries. Its approach may matter more as AI agents increase demand for systems that can retrieve and analyse business data quickly, including in natural language.
The company grew out of a decade of research at the Technical University of Munich and is based on the Umbra project, designed to make efficient use of modern hardware. CedarDB is compatible with PostgreSQL to reduce the effort of switching; it launched free and enterprise on-premise versions alongside a €5.3 million seed round in May last year. It has now announced limited availability of its managed cloud service, initially serving users with a 14-person team, and sees Databricks’ similar push towards unified data as evidence of demand.
- CedarDB combines transaction processing and analytics in one database.
- Its managed cloud service is now in limited availability.
- PostgreSQL compatibility is intended to make switching easier.
New here? Start with this
Business databases handle day-to-day tasks such as recording orders, while separate analytical systems are often used to examine large amounts of data. Running both kinds of work on one system could reduce the need to move information between them and help businesses get answers more quickly.
CedarDB is a Munich-based company whose database grew out of research at the Technical University of Munich. It is designed to work with PostgreSQL, a widely used database, which may make it easier for organisations to try it with existing software.
The company is also offering a managed cloud service, where it operates the system for customers. The broader question is how businesses will store and query their data as AI agents—software that can carry out tasks—are used to find information and analyse it.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
A single database for transactions and analytics could reduce the complexity, delay and cost of moving data between separate systems. If CedarDB can deliver strong performance and PostgreSQL compatibility, businesses may be able to build applications and AI agents that work with current data more simply, while keeping deployment on-premise or in the cloud to suit their needs.
The case against
Combining operational and analytical workloads can create difficult trade-offs: each may compete for resources, and a newer system may lack the maturity, reliability or ecosystem of established databases and data platforms. For businesses, adopting a small startup’s managed service could also add migration and continuity risks, while separate specialised systems may offer more predictable performance and room to scale.