Company intelligence
MinervaDB
The World's Most Trusted Full-Stack Database Infrastructure Partner — Architecture, Engineering, Operations & Analytics
About MinervaDB
MinervaDB is the world's leading vendor-neutral, full-stack database infrastructure firm — the partner enterprises trust when their data platform cannot afford to fail. Founded in 2015 and serving 800+ clients across 40+ countries, MinervaDB delivers end-to-end ownership of the entire database lifecycle: Architecture, Engineering, Operations, and Analytics. Our principal engineers design systems built to perform at 100x scale, operate them around the clock to explicit SLOs, and migrate mission-critical estates without drama. We carry production-grade depth across 13+ engines: PostgreSQL, MySQL, MariaDB, Microsoft SQL Server, MongoDB, SAP HANA, ClickHouse, Trino, Apache Cassandra, Redis, Valkey, and Milvus — and every major cloud DBaaS across AWS, Azure, and GCP. What sets us apart: singular accountability. The engineers who architect your platform are the same engineers who carry operational responsibility for its performance. No finger-pointing. No junior staff learning on your dime. No vendor agenda. 24x7 global coverage. 99.999% availability targets. $900M+ in documented client cost savings. Performance improvements of 200-5,000x delivered on mission-critical workloads. MinervaDB: the database infrastructure your business can rely on — engineered, measured, and operated to a standard the rest of your stack depends on.
Verified activity
Signals from MinervaDB
4 published signals
Research & Knowledge
MinervaDB spent the week on ClickHouse 26.8 to explore real statistics inside the database, including t-digest quantiles, Welch and Mann-Whitney tests, correlation, and stochastic regression.
Reported by Shiv Iyer
Products & Services
MinervaDB released Valkey 9.1 Performance, detailing what changed inside the engine.
Reported by Shiv Iyer
Products & Services
MinervaDB released MySQL 26.7.0, which shipped the Change Stream Applier, an opt-in replacement for the multithreaded replica applier.
Reported by Shiv Iyer