AI that works: Apify + TwelveLabs

AI that works is a monthly meetup for anyone building with AI in Austin: developers, founders, operators, go-to-market (GTM) teams, and the AI-curious. Every session: live demos from people shipping real things, a hands-on build, and honest conversation about what's actually producing results. We don't care if you're writing code or mostly prompting. If you're putting AI to work on something real, you belong here. This month We'll build a competitor ad-spy tool that turns hours of manual scrolling through the Meta Ad Library into a paste-one-URL workflow. You give it a competitor's Facebook page, and it pulls their video ads, ranks them by the spend and longevity signals that reveal a likely winner, and has AI watch each one to reverse-engineer the creative strategy: hook, offer, proof, and call to action (CTA). If you run paid social, you'll leave knowing how to surface a competitor's winning ads and the structure behind them. If you build with AI, you'll leave with the full stack: an Apify Actor that orchestrates a scraper, a scoring model, and four different AI models into one tool. How the evening works We start with why studying competitor ads by hand doesn't work: you scroll the Ad Library, watch dozens of videos one at a time, screenshot hooks into a doc you never reopen, and you're still guessing which ads are actually winning. Then we build the tool that does it for you. First, the data layer. We wrap apify/facebook-ads-scraper inside a single new Actor that pulls up to 100 ads per page with public metadata - active status, run duration, placements, variants, and spend signals - then scores and ranks them so the high-conviction ads float to the top. Next, the analysis layer. We pass the top-ranked videos to TwelveLabs: Pegasus 1.5 returns a structured creative blueprint (hook, audience, problem framing, offer, proof, CTA, pacing, and scene-by-scene notes), and Marengo 3.0 indexes each video so you can query it later. Gemini then synthesizes a cross-competitor report on the patterns across hooks, visual styles, offers, and CTAs. Finally, the interrogation layer. We wire up a split-screen video-and-chat view powered by Claude Sonnet, ask it "show me the hook" or "where's the CTA?", and click any returned timestamp to seek the video straight to that frame. You'll leave with a working composite Actor in your Apify account and a repeatable pattern for orchestrating multiple models behind one clean input and one useful output. Bring your laptop. Agenda What you'll need for the hands-on portion Don't have all of this? No worries, you can follow along with someone at the event who does. The concepts apply regardless of your stack. About the host Apify is the platform behind thousands of ready-to-use tools that give AI agents and apps access to web data, automation, and real-world information. But this meetup isn't a product pitch, it's a community. See what people are building, learn something new, and meet the people doing the work. 👾 Join the conversation on Discord. Explore Apify at apify.com. We will use your information to contact you about the event. For more information, see our Privacy Policy.