Building the future of storytelling.
The models are finally good enough. The gap between AI-generated content and actual production quality is closing. Our job now is to bridge it completely. Not AI slop. Real stories that hold up next to anything made the traditional way.
We're shifting from one-shot creations to becoming a full platform for content creators. An AI-native video editor that gives creators real control. The technology caught up to the vision. Now we build the tools that let people use it.
Leadership changed. Someone had to make sure things kept moving. The priority was simple: stabilize the team, keep the momentum, make sure nothing falls through the cracks.
Remote team across timezones, sprint planning, product decisions. The goal was smoothness — no friction, no gaps. Keep shipping, keep the team aligned, keep the company on track.
Hiring is harder than building product. We flew candidates in, ran trials, learned that finding people who actually care is the real challenge. Built a culture where the team stays because they want to, not because they have to.
The ones who stuck around became family. Late night debugging sessions, shared wins, shared frustrations. A startup is just people believing in the same thing long enough to make it real.
The hardest part wasn't the technology. It was understanding what makes a story actually work. We partnered with creators, watched hours of content, studied what separates forgettable from unforgettable.
Pacing. Emotional beats. Visual rhythm. The stuff film schools teach in years, we had to figure out and encode into AI systems. Every conversation with a creator became a lesson. Every piece of feedback became a feature.
The goal shifted from “generate videos” to “generate stories worth watching.” That meant building AI that understood creativity, not just followed prompts.
Growing users meant growing infrastructure. Auto-scaling, database architecture, API reliability. The unsexy work that makes everything else possible. When the product works, nobody notices the backend. When it breaks, everyone does.
2024 was the year image models got good. Flux came out of Black Forest Labs and suddenly AI-generated images were actually usable. We built around that. Then 2025 hit and everything shifted to video. Sora, Veo, Kling. Models that could generate actual motion, actual scenes. The technology finally caught up to what we were trying to build.
I built the orchestration layer that ties it all together. LangGraph workflows coordinating specialized AI agents. PostgreSQL state persistence so sessions survive server restarts. Real-time WebSocket streaming so users see progress as it happens. The plumbing that turns bleeding-edge models into a product people can actually use.
$2M raised. Coming out of HF0 with real traction meant investors were paying attention. The fundraise opened doors for hiring, but also raised the stakes on everything we built.
Packed up and moved from New York. The startup scene, the energy, the people who understand what it takes to build something from scratch. SF felt like the right place to be.
Twelve weeks in a mansion with the top 10 teams. They said it would be the most productive weeks of our lives. They were right.
We pivoted the entire product. From children's books to AI video generation. Rebuilt everything. New backend, new infrastructure, new vision. Ship it, fix it later, move fast. Demo day was March 19th. Story.com was born.
Nobody expected StoryBird to grow like it did. Text-based AI content meant SEO worked in our favor. Every story generated was indexed, discoverable, bringing in organic traffic we never paid for.
The revenue and traffic numbers for a product this simple surprised everyone, including us. That's what caught HF0's attention. Not a fancy pitch deck, just real traction that didn't make sense on paper.
Amazon publishing integration changed everything. Suddenly users weren't just generating stories, they were publishing actual books. The product had found something real.
We kept shipping features, optimizing the pipeline, watching the numbers climb. A small team building something people actually wanted. No playbook, just iteration and paying attention to what worked.
Three person team. Remote. I was in New York building the backend from scratch. The product was StoryBird.ai, an AI-powered children's book generator. Simple concept, but it worked.
My job was to make things work. Backend systems, ML pipelines, infrastructure, whatever needed building. That never changed, even as everything else did.