A company migrated trillions of rows of IoT vehicle GPS data from Parquet to self-hosted ClickHouse, discovering superior compression through optimized schema design and S2 geospatial indexing. They achieved better compression ratios than their previous Parquet setup by carefully choosing sort orders and encoding strategies to balance compression efficiency with query performance.
Bold.org built an internal AI agent starting in January 2026 that investigates production issues via ClickHouse queries and integrates with Slack, Linear, PostHog, and Grafana. The agent reduced investigation time from hours to zero-to-one hour and spread across engineering, product, support, operations, marketing, and growth teams through visible shared usage rather than formal launches.
ClickHouse uses a proprietary storage format developed before open standards like Parquet existed, requiring data duplication from other systems. The authors question whether a modern real-time analytics engine built natively on open formats like Parquet could match ClickHouse's performance, and introduce Pivot, an open-source database that achieves comparable or better speed while using standardized formats.