Real-time data, real-world impact: Building a high high-performance feature store

Arpeely operates in the demanding AdTech environment, processing millions of ad requests per second with sub-300ms latency. This session explores their highly configurable, real-time feature store. We'll examine how it enables data scientists to define, modify, and test feature configurations by combining data from diverse sources. Learn about the architectural challenges, trade-offs, and innovative solutions behind Arpeely's scalable, low-latency system that drives their real-time advertising success.
Speaker
Real-time data, real-world impact: Building a high high-performance feature store

Arpeely operates in the demanding AdTech environment, processing millions of ad requests per second with sub-300ms latency. This session explores their highly configurable, real-time feature store. We'll examine how it enables data scientists to define, modify, and test feature configurations by combining data from diverse sources. Learn about the architectural challenges, trade-offs, and innovative solutions behind Arpeely's scalable, low-latency system that drives their real-time advertising success.