Fringeon
Figure and metrology control from interferometer to machine.
Precision Optics
Foceon
Autonomous operations layer for precision-optics fabs — a fleet of AI agents and an optics digital twin that figure every surface to spec.
Overview
What Foceon does.
Foceon runs the precision-optics fab: metrology and figure control, autonomous grind and polish, thin-film coating, defect inspection, robotic cleanroom handling and the optics digital twin.
Every lens and mirror deployment produces labelled surface and process traces that train the next control model.
| Detail | Value |
|---|---|
| Stage | Seed |
| Category | Physical AI |
| Year | 2026 |
| Website | foceon.com |
| Headquarters | Not disclosed |
| Raised | $1M |
| Investor | 10xC |
| Model | Per-cell software subscription with an outcome component on rework saved. |
| Measured by | Figure conformance, rework rate, cycle time per surface, cells expanded per customer. |
Products
One loop, module by module.
Every deployment on foceon.com writes labelled traces back into the model, so the system generalises across plants, alloys, grades and geometries.
Polira
Autonomous grind and polish with closed-loop surface control.
Coatix
Thin-film coating process control.
Scryon
Defect and surface-quality inspection.
Vitreon
Optics digital twin for process and ramp planning.
Cleora
Robotic cleanroom handling.
Why we invested
The thesis behind Foceon.
Precision optics is a craft business sold at the limit of what people can figure by hand. Fabs depend on the touch of a few master polishers and on rework when a surface drifts off spec.
Foceon puts the figure in software. The buyer is an optics fab paying in hard currency; the wedge is one polishing or metrology step, and rework saved is the number that opens the next module.
Who is building it
Co-Founder & CEO
Co-Founder & CEOOwns the fab relationships, the pilot programme and the commercial model.
Co-Founder & CTO
Co-Founder & CTOOwns the figure models, the edge runtime and the write-back into polish and coat control.
Building something similar?
Building in Physical AI? We want to hear about it.