Let's Build: Jockey Video Intelligence

​You may remember TwelveLabs from Denver's first multimodal AI hackathon in 2024. They're back, and they're bringing Jockey.

​Jockey started as one of the first multimodal video agents ever built. Now it is improved and available through an API. Point it at a corpus of video and ask a question. It can reason across every frame, clip, and spoken word. Cited clips, timestamps, and streaming URLs are returned in a single response.

We'll start with a quick overview of what the new models can do. Then we build.

By the end of the session you'll create an application that can search, analyze, and reason over video using natural language.

5:30 - Pizza and drinks
6:00 - Overview and hands-on workshop
7:00 - Demos & socialize

Don't miss this. We're lucky to catch our friends while they're in Denver for CVPR.

​Bring a folder. Bring a question. Bring your own video that you've never been able to search before.

Bring your laptop. Let's build.