FIELD / 01
Machine learning & computer vision
We design, train and run our own models — vision systems that understand motion, people and 3D space. Built in-house, end to end, so we know exactly what they can and can't do.
Spill Studio
An independent AI product studio. We take machine learning and computer vision past the demo — into products people rely on for real work.
The studio
Most AI never leaves the lab. Ours ships.
Spill Studio is a small team doing the whole job: the research, the infrastructure, and the product around it. We pick hard perception problems — motion, geometry, people — and turn them into tools professionals trust with production work. Our first product, mimem.ai, converts plain video into 3D animation data.
FIELD / 02
Product craft
Research is half the job. We build and run everything around the model — pipelines, interfaces, support — the unglamorous parts that make it dependable.
FIELD / 03
Go-to-market
Shipping isn't the finish line. Positioning, pricing, onboarding, support — we take our products to market ourselves, and every user conversation feeds the next release.
Shipped — field proof
mimem.ai — motion capture without the suit.
See it workmimem turns regular video into 3D animation data. No markers, no suits, no capture stage — footage from standard cameras is enough. Game studios, animators, VFX teams and animation schools use it for everything from a single move to full-length sequences.
5+
Cameras, one take
Minutes
From footage to animation
0
Frames invented by AI
How we work
Trust is a feature. We build for it.
-
Faithful over flashy
A tool that guesses is a liability in production. We'd rather tell you what our tech can't do than sell you something it can't deliver.
-
End to end, in-house
Research, models, infrastructure, product, support — one team. When something breaks, the person answering your email can read the logs.
-
Trade-offs, out loud
Every technical choice costs something. We put those costs in the documentation, not in the fine print.
Contact
Something worth spilling?
We're open to product partnerships, investment conversations, and hard perception problems.