Case Study: Tubi achieves real-time personalized streaming and higher ad revenue with Akka

Tubi cuts personalization infrastructure from 150,000 to 22,000 cores with Akka

Tubi, a leading ad-supported streaming service, faced the challenge of personalizing content recommendations in real time across a massive catalog of over 300,000 titles to boost viewer engagement and advertising revenue. This required a platform capable of processing live data streams, scaling automatically, and ensuring resilience without diverting engineering resources to infrastructure management. They turned to the Akka SDK to address this.

By implementing Akka, Tubi’s engineering team could focus on improving personalization logic instead of building infrastructure. The solution delivered significant results: a reduction in the core count for the engine from 150,000 to 22,000, a 40% lower cost per session, and the ability to process 5 billion tokens per second for real-time personalization. This partnership with Akka led to increased development velocity and, ultimately, hyper-personalized experiences that grew Tubi's addressable advertising revenue.

Tubi