PhiloticTM

Specialized intelligence, on open foundations.

We build domain models for work that general models get almost right: finding what’s wrong in a photo, naming it, placing it, and grading it. Then we publish how they do, on held-out data, with the misses.

Explore the modelsTalk to us
A bucking horse throws dust in an arena as its rider lands.

Foundation models

General purpose

Open-weight models with commercial licenses, for reasoning over images and language. You choose the one you build on.

Philotic domain models

Specialist

Specialists for the calls experts make, published on held-out data, and made yours with your own people’s decisions.

Latest research.

A technician checking an engine under an open hood.

Where we work.

Visual inspection

Finding every damaged or out-of-spec area in a whole image, and saying where it is.

Grading and assessment

Rating what was found on the scale your people already use, learned from their own grades.

Motion and pose

Measuring how a body moves through a clip, so a score can be traced back to what happened.

Adaptation

Small adapters that carry a domain model to one customer’s data and conventions, owned by that customer.

Evaluation

Held-out test sets, paired comparisons and bars agreed in advance, so a model goes live only past the bar you set.

Distillation

Making models small and fast enough to run on modest hardware inside your boundary.

Build the model your work needs.

Tell us what your people inspect, grade or decide every day. We’ll test candidate foundations on your task and show you the results before anything goes live.

Talk to usRead the research