Condé Nast cuts video search from 250 minutes to under 2

Condé Nast's editorial teams were spending an average of 250 minutes per content discovery task, manually scrubbing through more than 140,000 videos using only titles and descriptions. The core problem was structural: existing search tools couldn't look inside video content, so teams depended on institutional knowledge to locate assets, creating single points of failure when specific individuals were unavailable. Underutilized content also sat in the archive undiscoverable because no keyword in a title or description connected it to the queries editors were actually running.

To solve this, Condé Nast partnered with the AWS Generative AI Innovation Center to build a solution on Amazon Bedrock and Amazon OpenSearch Service. They selected the TwelveLabs Marengo embedding model for its ability to jointly encode visual, audio, and transcript signals. The design separates an asynchronous ingestion pipeline from a synchronous serving tier, so search stays available while the pipeline reprocesses content. The solution has been running in production for six months.

The results hinge on whether the productivity gains measured in the May 2026 benchmarking workshop hold as the library grows and editorial workflows evolve. For Condé Nast, the estimated $800,000 in annual operational savings and faster response to advertiser requests suggest the investment is paying off, but the broader test for other media organizations will be whether similar decoupled architectures can deliver comparable gains without extensive customization.