Case Study: Poolside scales AI development with ClickHouse Cloud

How Poolside is using ClickHouse to build next-gen AI for software development

Poolside, the AI company focused on helping developers build software more productively, was struggling to keep pace as its data operations and training workloads scaled rapidly. With billions of documents, more than 800,000 real-world codebases, and clusters growing to 10,000 GPUs, its existing ad hoc data pipelines could no longer support the speed and scale needed for iterative AI development. Poolside turned to ClickHouse Cloud to address these challenges.

Using ClickHouse Cloud, along with ClickPipes and standardized Spark pipelines, Poolside streamlined data ingestion and analytics across its workflows. The platform now supports around 50 terabytes of compressed data and enables queries across billions of records in seconds, helping the team move faster from idea to dataset, run more experiments, and refine models more efficiently. By adopting ClickHouse, Poolside eliminated major data bottlenecks and gained the performance and simplicity needed to keep scaling its AI research.

Poolside

Eiso Kant

Chief Technology Officer