Markets
SereneDB Open-Sourced Krummelanke, a Postgres-Compatible Engine That Fuses Elasticsearch-Style Search With ClickHouse-Style Analytics for Agent Query Load
Berlin-based SereneDB released Krummelanke on September 22 under Apache 2.0, pitched as the first production-ready database designed for fleets of AI agents rather than human dashboard users, with roles and permissions enforced for agents at the data layer and indexing of data where it already lives instead of copy-first ETL. CTO Andrey Abramov claims a design-partner workload that cost about $2,000 a month in infrastructure now runs at roughly a tenth of that, with configurations and methodology published on GitHub. The repo shows 780 stars as of September 22; the company raised a $2.1M pre-seed from HTGF and Entourage in December 2025, so treat the benchmark claims as vendor-run until someone reproduces them.
Source
↳ Follow the thread