QuestDB

Open-source time-series database built for demanding workloads that require ultra-low latency and high ingestion throughput. Delivers a multi-tier storage engine with SIMD-accelerated SQL queries, write-ahead logging for durability, and native support for Parquet and standard SQL.

Trusted by enterprises such as Airbus, B3, Mizuho, and OKX for capital markets and real-time applications involving billions of data points. QuestDB differentiates from TimescaleDB and InfluxDB with its time-series–optimized SQL: ASOF JOIN for time-bounded event alignment, SAMPLE BY for resampling and gap filling, and multi-dimensional arrays for order-book analytics. Postgres wire protocol compatibility and Apache Parquet output keep data portable across tools and AI frameworks.

Key capabilities:

• SIMD-optimized, multi-threaded SQL engine with time-series extensions such as ASOF JOIN and SAMPLE BY
• Write-ahead logging for durable, high-throughput ingestion with time-partitioned columnar storage
• Native Parquet support with tiered storage for local and object storage, queryable through a single SQL interface
• Postgres wire protocol compatibility for drop-in integration with existing clients
• Streaming materialized views and multi-dimensional arrays for real-time rollups and order-book analysis

Common workloads: capital markets tick data, trades, order books, and OHLC at nanosecond resolution; IoT and sensor streams; observability and metrics dashboards. Integrates with Kafka, Telegraf, Grafana, Pandas, and Polars. Developers connect via Postgres drivers or HTTP, and can query Parquet files directly through QuestDB SQL.