Case Study: Poolside achieves faster LLM training and AI model development with Redpanda BYOC

How poolside uses Redpanda BYOC to stay ahead in the competitive world of AI

Poolside, a software and technology company building AI-driven software development tools, needed a better way to connect real-time data pipelines for training large language models (LLMs). After initially using the Apache Kafka ecosystem, the team found it was not well suited to the throughput, scale, and cloud-native needs of its AI workloads, so they turned to Redpanda Data and Redpanda Cloud BYOC.

Redpanda Data implemented Redpanda BYOC so Poolside could keep full control of its infrastructure and data while offloading cluster management. The solution provided Kafka API compatibility, better visibility, lower operational overhead, and GDPR-friendly data sovereignty in AWS-backed VPCs. As a result, Poolside reduced model training time from weeks to a day, moved data faster, and freed engineers to focus more on building new AI features and improving model quality.

Poolside