Brennan Erbz

CEO & Co-Founder

Flik

Brennan Erbz is the CEO and Co-Founder of Flik, a generative computer that produces complete creative work across text, image, video, and audio. The multi-agent, multi-modal workspace has a 50,000 person wait list. Erbz has a habit of building things before the market knows it needs them. He was an early machine learning engineer at Snap before its IPO, then founded Hashletes - a blockchain trading platform that secured an official NFL license and Khosla Ventures backing before being acquired by JDS Sports. A Y Combinator alum (W20), he spent the years that followed going deep on the AI infrastructure and multi-agent architecture that became Flik. The engineering team he's assembled draws from AI researchers at UPenn and Cornell with published work in multi-agent systems, alongside creative technologists from NYU Tisch and USC who build at the code level. Together they represent over 36 months of compounding technical work that no competitor has come close to replicating.

Sessions (1)

Beyond Prompts: Building a Multi-Agent Creative Computer That Orchestrates 5+ AI Models in Real-Time

3:20 PM·Track 1 · Room 2010

Flik is a production multi-agent system that generates complete work, not pieces. A 20-minute TV episode with screenplay, consistent characters, 50+ animated scenes with dialogue and lip sync, original score, and final edit. 500 product shots from a single image. Three months of social content in a day. What used to take 15 vendors and three months now happens in days. This is a demonstration of HOW it works. We'll show how Flik orchestrates Claude, Gemini, Nano, Seedance, and Eleven Labs in a single workspace to handle multiple creative functions simultaneously. We'll show: -The multi-agent workspace in action - A live walkthrough showing how different AI models work together to generate complete creative work. We'll demonstrate one end-to-end workflow from prompt to finished output. -How models coordinate across modalities - The architecture that maintains consistency across text, image, video, and audio generation in a single project -Production safety built into the system - How the platform blocks likeness misuse and IP violations before generation happens -Real customer examples - Quick examples from our 50K waitlist showing complete e-commerce catalogs, training programs, and campaigns generated in days instead of months This is a live demonstration of a production multi-agent system shipping to paying customers.

Generative Mediaintermediatetalk