Case Study: PBS enhances digital search with Amazon Bedrock from AWS Elemental

PBS enhances search for 700,000 assets in less than six months with AWS Elemental

PBS sought to improve how viewers discover content on its PBS App and PBS LearningMedia platforms, needing a way to enhance search beyond basic keyword matching. To address this challenge, PBS collaborated with AWS and utilized Amazon Bedrock to build a new generative AI-powered search capability.

The solution implemented with AWS involved using Anthropic's Claude Sonnet model on Amazon Bedrock to automatically analyze and apply detailed metadata tags to PBS's entire library of over 700,000 assets. This project, completed in less than six months, enabled viewers to search by nuanced criteria like themes and moods. The result was a vastly improved search experience delivered efficiently and at a fraction of the cost of manual tagging, allowing PBS to successfully scale the solution into production.

PBS