Where AI meets
the products behind
everyday life.

We believe there is a smarter, more efficient way to create the products that reach every person on the planet.
One where intelligence does the heavy lifting.
So the best products reach more people, faster.

That's what we're building.

Matías Muchnick
Co-Founder & CEO

What if 80% of what happens in a lab could happen on
a computer instead?

The trials, the errors, the conditions nobody anticipated.
All of it, computationally. Every cycle compounds, making
the next one sharper. Your team starts discovering
combinations nobody would ever think to try. The product
that comes out is the one you imagined. Often better.

That's the shift we're building for every product people enjoy, every day.

In good company.

Partners

Investors

Our story

From the lab to
enterprise AI.

Moment zero

Product formulation is often thought of as a chemistry problem. We like to think of it as a data problem.

Every flavor you've ever liked, every texture that made something feel right, every reformulation that had to hit a new cost target without anyone noticing. That's a multi-objective, combinatorial problem of staggering complexity. Cost, sensory, nutrition, manufacturability, compliance. All at once. Change one thing, and something else breaks.

For decades, the answer was talent and time. Brilliant food scientists. Long lab cycles. Accumulated knowledge passed down through teams. It often produced remarkable things. But it also produced a ceiling. More hours didn't solve it. More trials didn't guarantee success. And no one was building the tool to break it.

Born in the lab

In 2015, we saw it up close. In the labs, with the scientists, inside the problem. That's when we founded NotCo AI.

Because much like LLMs recognized that knowledge work could be transformed by foundational models, we believed AI could do the same for the physical world, starting with the products people consume every day. Not just speed things up. Change what was possible.

So we didn't build from the outside in. We spent the first years building AI by doing the work: formulating real products, running experiments under real manufacturing constraints, with real scientists, facing the same problems our clients face today. That's how we generated the proprietary data no one else had. Not scraped. Not synthetic. Earned.