Venture Building
Building What People Need
TLDR: I joined Entrepreneur First, launched and killed multiple ideas, rebuilt the team from scratch multiple times. Eventually, we built an agentic AI pipeline that cut content review by 70% for global brands. The product was acquired.
The biggest lesson from my previous company (Brainamics) wasn't technical. It was that even brilliant technology doesn't matter if there's no real market need. I'd fallen in love with the solution. So when I joined Entrepreneur First, I forced myself to do the opposite: prototype fast, talk to users, and kill anything that didn't solve a real problem. I launched and invalidated multiple concepts in quick succession, from agentic knowledge extraction via knowledge graphs to real-time video recording copilots, before finding something that stuck.
The answer was content review workflows. Brands were spending weeks manually checking whether influencer content met their guidelines before it could go live. I built Brandwagon to attack that bottleneck: an agentic AI pipeline handling video ingestion, multi-criteria compliance scanning, automated feedback generation, and delivery, running on a distributed, event-driven platform I architected for high-throughput inference. I invested heavily in CI/CD so my team could ship client-requested features to production in under 30 minutes.
Manual review work dropped by roughly 70%, turnaround went from weeks to days, and we served global brands including Opera, Tractive, and Gowish. I raised six-figure investment, and the company was acquired. I stepped away to return to the intersection of AI and science, which is where I've always done my most motivated work.
Tools & Methods
- 0-to-1 product discovery
- Stakeholder interviews
- Rapid prototyping
- Multimodal AI
- Distributed workers
- Python
- Redis
- Enterprise onboarding
Next step