According to a recent LinkedIn post from TwelveLabs, the company and Pixeltable built a video recommendation engine designed to identify content within long-form videos rather than relying only on titles, tags or upload dates. The post frames the system as a response to discovery challenges across podcasts, interviews, lectures and documentaries, where larger catalogs can make it harder for viewers to find relevant material.
The described product could support personalized “For You” feeds, creator pages, cross-catalog “Up Next” recommendations and explanations for why a video was selected. TwelveLabs’ Marengo and Pegasus tools are presented as analyzing scenes and adding structured context, while Pixeltable is described as providing the infrastructure for tunable discovery features.
For investors, the collaboration suggests a potential route to monetize video-understanding capabilities through recommendation and engagement tools for platforms with extensive libraries. If implemented at scale, better discovery could increase viewing time and improve catalog utilization, although the post does not provide customer details, financial terms, adoption metrics or evidence of commercial deployment.


