November 2025 Summaries

QuestDB 9.2 is an open-source time-series database designed for high-demand workloads, offering ultra-low latency, high ingestion throughput, and a multi-tier storage engine with native support for Parquet and SQL. The release introduces several enhancements, most notably the native DECIMAL type for exact arithmetic, which is crucial for financial operations requiring high precision, and a Dense ASOF JOIN algorithm that optimizes temporal joins for workloads with distant matches. With symbol capacity auto-scaling now enabled by default, the database handles high cardinality datasets more efficiently. The update also includes improvements in engine robustness, such as resilient asynchronous jobs and safer deduplication, alongside performance enhancements in ASOF JOIN strategies. These features are aimed at enhancing precision-heavy workloads and complex temporal analytics, making QuestDB an attractive choice for capital markets, industrial telemetry, and other applications where time and precision are critical.

QuestDB is an open-source time-series database optimized for high-frequency financial data processing, offering ultra-low latency and high ingestion rates, making it suitable for demanding environments like trading floors. It supports Parquet and SQL, ensuring data portability and AI readiness without vendor lock-in. The text outlines a tutorial for building an FX liquidity stress analysis pipeline using QuestDB and Python, aimed at predicting stress events in currency markets characterized by widened bid-ask spreads and reduced market depth during uncertain times. It employs QuestDB for streaming and aggregating high-frequency data, feature engineering, and training an XGBoost model for early stress detection. The pipeline involves data ingestion into QuestDB, feature calculation using SQL and Python, data export to Parquet for machine learning, and a demonstration of prediction on live data. Despite modest model performance, the prototype highlights QuestDB's capabilities in handling dense financial data, enabling quick experimentation and real-time analysis for applications in risk monitoring and market research.