Condé Nast, working with the AWS Generative AI Innovation Center, built a multimodal video discovery solution on Amazon Bedrock and Amazon OpenSearch Service that cut search time across a 140,000+ video library from 250 minutes to under 2 minutes per task.

1 min read

Condé Nast Cuts Video Discovery From 250 Minutes to Under 2 With Amazon Bedrock

What happened

FAQ

What did Condé Nast build with AWS?

A multimodal video discovery solution built on Amazon Bedrock and Amazon OpenSearch Service, developed with the AWS Generative AI Innovation Center, to speed up finding clips inside a library of more than 140,000 videos.

How did the search numbers change after deployment?

According to the source, the editorial team averaged 250 minutes per search task using titles and descriptions alone; after the solution, that dropped to under 2 minutes per task.

Can MENA organizations use this model?

The pattern applies to any organization with a large video archive, such as broadcasters, government entities, and publishers, because the architectural components (Bedrock and OpenSearch) are available as managed cloud services and can be adapted to an organization's own data.

What is the difference between text search and multimodal video search?

Text search relies only on titles and descriptions, which may not reflect actual video content. Multimodal search analyzes the video content itself alongside text, improving result accuracy and reducing the time needed to find the right clip.

Source: AWS Machine Learning

AI-assisted summary of the source above. High-confidence stories publish automatically; others wait for review. Facts that are not in the source are not added.