Case Study: PBS delivers personalized viewer recommendations with ClearScale and Amazon Personalize

PBS delivers personalized recommendations for 100M viewers with ClearScale

The Public Broadcasting Service (PBS) wanted to build a sophisticated smart recommendation engine to offer a truly personalized experience to its millions of viewers. To achieve this, PBS partnered with ClearScale to leverage AI/ML expertise and develop a solution on the AWS platform, specifically using Amazon Personalize.

ClearScale implemented a comprehensive MLOps platform for PBS, which included data pipelines and four distinct machine learning models for recommendations. The solution yielded a "Precision at 10" metric of 0.0706, meaning with every 10 titles recommended, at least one will be favored by the user with 71% probability. This significantly enhanced viewer personalization and provided PBS with a scalable system to improve engagement.

Public Broadcasting Service (pbs)