FUTURE
Teaching applied AI to see and understand.
We begin in galleries and museums from a place of genuine passion — and because training in a world where every object is N-of-1 gives us enormous efficacy everywhere else.
01 — Why art first
The hardest possible training ground.
Rare objects are the most difficult category of visual problem there is. No two are alike. Condition, provenance, and setting change what a piece is worth. Value is contextual, contested, and moves over time — and there is rarely a second example to check against.
Most vision systems are trained where repetition is abundant: thousands of identical parts, identical frames, identical scans. We chose the opposite. A model that can recognize, situate, and value an object it has never seen before generalizes downward into domains with more repetition, not less.
Curator and Assessor put that model in the hands of people who work among the objects themselves — running on the phone or tablet already in their pocket, and on a third-party 3D camera when a piece warrants it. The gallery and the museum are our first field deployments, not our ceiling.
02 — The model
Built to read the world as it is.
Four complex dimensions most models ignore entirely. Together they form a unique and durable advantage that compounds with every object the system sees.
Physical
Material, condition, density, scale, and the way light falls on a surface — read from a camera, not a spreadsheet.
Temporal
How an object changes: ageing, restoration, cycles of wear, and the shape of a price over decades.
Evolutionary
How categories, makers, and taste develop — lineage, influence, and where a body of work sits within it.
Contextual
Setting, ownership, provenance, and comparables. The same object means something different in a different room.
03 — What's next
Four verticals where the same capability applies.
Agriculture
- Smart weeding, planting, packing companies
- QA during harvest and following storage
- Spot problem fields in advance
Mining
- On- and in-ground tactical mineral discovery
- Smart safety flagging
- Spot impurities at microscopic scale
Medical
- Dermatology and scan interpretation
- Home monitoring
- Empower patients to understand their diagnoses
Industry
- QA in production and manufacturing
- Instant, accurate inventory
- On-floor supervisory technology
Investors
Back the laboratory.
A marketplace that funds a research programme. See the thesis, the margin structure, and the roadmap beyond art.
Request the data roomPartners
Work with us early.
If you operate in agriculture, mining, medical imaging, or manufacturing and want to pilot the model, we'd like to hear from you.
Start a conversation