Before building Stylus, I spent a year as product lead at an AI platform company, designing core AI workflows in close collaboration with professional film and video production teams. What I kept finding: the gap between AI tools and real production workflows was enormous. Most AI efficiency gains weren't being systematically adopted by professional creative teams. I wanted to build something that could actually change that.
The content that truly moves people comes from creative instinct, accumulated taste, and a deep understanding of the audience. That's something AI will never replicate.
AI takes over the process work so the team's creative energy goes where it belongs — into the work itself.
With AI-powered visualization, a professional team's creative vision can translate more completely into the final product — fewer compromises between what was imagined and what gets made.
Stylus — the oldest tool in a creator's hand. From Leonardo sketching compositions with a silverpoint on wooden boards, to today's directors storyboarding in Procreate with an Apple Pencil — every generation has solved the same problem: getting what's in a creator's mind onto a surface others can see and act on.
So we named our product Stylus — an AI-driven creative tool that keeps creative decisions flowing precisely between every member of the team, without losing anything in the handoff.
In the age of AI, production capability is available to everyone. What's truly rare — is your style.
Scripts, references, mood boards, AI tools — scattered across different apps and platforms. More tools than ever, but no single place where the project actually lives.
The director's vision passes through conversations, documents, storyboards, and briefs before it reaches the DP, art director, and producer. By then it's already changed. Misalignments don't surface until you're on set — and every reshoot costs real money.
A professional team's most valuable output is creative judgment. But most of the day goes to alignment meetings, tool-switching, and chasing down assets — directly taking away from the actual work.
The feeling a director has in their head — there's finally a tool that can receive it and turn it into something visual everyone can see.
Reference images aren't just mood board material anymore. They can become direct inputs for AI generation — turning into keyframes, storyboards, and video previews.
The visual style, tone, and atmosphere locked in for this project can become the foundation for the next one — compounding over time.
You don't need to change how you work. Just bring everything into one place — and let the information mean something.
Upload a brief or write from scratch. Agent auto-extracts scenes, characters, and tone keywords — and helps you draft and revise.
A free-form visual board. Drop in reference images, screenshots, sketches, anything. Or generate the visual feeling you have in mind directly with AI.
AI reads both your script and your visual direction, then generates a first-draft shot list. Every shot's framing, movement, and tone has a source.
Based on everything it's read across your project, the Agent generates keyframes, storyboards, and video previews — reviewed internally and with the client before you shoot.
Once your script, visual references, and shot list live in one place, AI reads the full project context — and something previously out of reach becomes possible.
Not random — generated from your Canvas visual language and shot parameters. Every frame has a reason.
String keyframes into a video. Let the client see roughly what it feels like before you shoot — and redirect on the spot if it's off.
Script, shot list, visual references, keyframes — organized and ready to send to the whole team. No hunting for assets again.
Based on all pre-production context, AI previews the entire film before you shoot. You've already "shot it once" — the actual shoot is just turning pre-viz into real footage.
After the shoot, AI matches the original project context to generate reshoot plans or AI replacement shots — anything that doesn't need to be reshot, won't be.
Alignment meetings and asset-hunting drop significantly. Every project runs faster.
Visual Assist surfaces direction problems before you shoot — not on set when fixing them is expensive.
Same team, more projects per month. Capacity gains translate directly into revenue.
Time saved, fewer reshoots, more revenue — it all adds up to better margins on every project.
One year of end-to-end product experience in the film AI space — user definition, workflow design, competitive positioning, and GTM strategy.
Senior AI Researcher at AMD, former core engineer at an AI infrastructure company. Deep background in AI systems and cloud engineering.
We're not looking for users. We're looking for partners who want to help build this right. Early teams will have a direct hand in shaping the product.