Physics and AI, applied to industry.

Neural Continuity Labs builds AI that understands the physics of industrial systems: models that see what happens inside a plant and anticipate what comes next.

Get in touch

Industrial systems are measured at a few points. What happens in between is left to experience. We build models that combine the physics of a system with the data it produces, to give engineers a complete and trustworthy picture.

Approach

Grounded in physics

Our models start from the laws that govern a system, rather than from data alone. This keeps them reliable where data is scarce.

Learned from the real world

They are trained on measurements from real systems, which are rich and noisy, so they capture what the equations leave out.

Built for operation

They run alongside the plant, fast enough to be useful, and report how much each estimate can be trusted.

Industries

Wherever a physical process runs at scale and is watched through a handful of instruments.

Energy
Manufacturing
Aerospace

Research

Models that do science need a test. We build the tests too.

NeurIPS 2026, Sim2Science workshop

KeplerBench

A benchmark for scientific discovery from observer-centered astronomical observations: Tycho Brahe's Mars record and synthetic worlds, scored on prediction, global law and local law.

Paper Code and data

About

Neural Continuity Labs grew out of doctoral research in AI and 3D reconstruction at Université Paris-Saclay.

The company is founded by Akash Malhotra and supported by Inria Startup Studio.

In the longer run, we want models that do science themselves: propose a hypothesis, test it against measurements, and improve. We start where it matters most today: industrial systems.

Contact

For pilot projects, partnerships or questions.

contact@neuralcontinuitylabs.com