QuestDB High-Performance Time-Series Database

QuestDB was founded in 2019 by Vlad Ilyushchenko and Nicolas Hourcard and lists its address in London, with team members in London and New York. It builds a time-series database for workloads that must ingest very high event volumes, compute on live data and keep long histories queryable, such as market data in capital markets and sensor data in industry.

QuestDB uses one SQL engine with PGWire compatibility and open formats such as Apache Parquet and Iceberg. The Apache-2.0 open-source edition runs on Docker, Kubernetes or bare metal, while QuestDB Enterprise adds high availability, replication, failover, multi-tier storage with cold partitions on object storage, RBAC, TLS and SSO. Enterprise runs on AWS, Azure and GCP or as BYOC in the customer's cloud account. Users include B3, HDFC Bank, Airbus, Mizuho, SIX and OKX.

• High-speed ingestion: QWP binary columnar protocol for writes and Arrow reads from one client.
• SQL on time series: time-series SQL extensions with PGWire compatibility.
• Cold storage: older partitions tier to object storage as Hive-partitioned Parquet while staying queryable.
• Enterprise HA: replication and restart-free failover triggered by a SQL command.
• Live views and notebooks: continuously updated window results and a Web Console with SQL and chart cells.
• Flexible deployment: open source, Enterprise on AWS, Azure, GCP marketplaces, or BYOC.

• Vietnamese securities companies storing and analysing market tick data for trading and risk dashboards.
• Banks monitoring payment and transaction metrics in real time.
• Energy and manufacturing firms keeping years of sensor history on low-cost object storage.

• Ingests up to 19 million rows per second per server, per the vendor.
• HDFC Bank serves more than 7,000 queries per second for 90 million users; BTG Pactual gets sub-10 ms API responses.
• Cold storage on S3 is about 85% cheaper on raw capacity than replicated block volumes at half a petabyte.
• Open formats (Parquet, Iceberg, SQL, PGWire) avoid lock-in.

Trading venues, brokers and crypto exchanges need tick-level analytics at ever higher volumes, while factories and energy assets add more sensors. Vietnam's securities market modernisation and the growth of digital brokerages and industrial IoT in Southeast Asia raise demand for fast, cost-efficient time-series storage.

Founded 2019; address 8 Devonshire Square, London, with team in London and New York. Co-founders Nicolas Hourcard (CEO) and Vlad Ilyushchenko (CTO). SOC 2 Type 2 compliant; listed on AWS and Azure Marketplace. 17.5k GitHub stars and 170+ contributors.