Now piloting: Catalogue Intelligence
The AI-native operating layer for cultural institutions.
Grow your capacity before you grow your staff.
Small museums, artist foundations, and archives carry enterprise-level complexity with 5 or 10 people. We give them the research, infrastructure, and operating capacity of a much larger institution while their experts keep the final word.
ASSERTIONCR-1974-0182
Claim → evidence → contradiction → human review
AI does the exhaustive work.Humans do the authoritative work.
SCHOLARSHIP
Scholarship that once cost millions. Now a fraction of the cost.
Infrastructure for cultural research, turning fragmented archives into source-backed, auditable knowledge, while scholars retain final authority.
$1M–$5M+
A traditional catalogue raisonné. Years of specialist labor.
50–80%
Target reduction in research-processing cost through automation of repetitive work.
5–10×
Target increase in scholarly research capacity, human review preserved.
Targets, to be validated through institutional pilots. Final scholarly decisions remain human.
01
Fragmented records enter one controlled research space.
PRACTICES
Built for the most demanding cultural work.
Six problems every cultural institution recognizes. Scholarship is our flagship and our proof. The others are where most staff hours disappear.
01pilot
Scholarship & Collections
Catalogue raisonné, provenance, archive intelligence, semantic and visual discovery.
02production
Digital Infrastructure
Cloud migration, server recovery, preservation, security, CMS and data migration.
ILLUSTRATIVE MODEL
Imagine 4,000 artworks.
32,000 hours
Traditional research at eight hours per work.
6,000–12,000
AI-assisted preparation at 1.5–3 expert hours per work.
20,000+ hours
Freed for attribution, interpretation, conservation, and writing.
Illustrative model. Actual savings depend on archive quality, digitization, and project complexity.
The same scholarship. A fraction of the hours. A fraction of the cost.
DEVELOPED WITH
Cultural institutions · Scholars and universities · Technology partners · Collectors and archival contributors
INSTITUTIONS
Built for every kind of cultural institution.
One institution rarely needs one thing. An archive that is impossible to search, a server nobody understands, and a catalogue that would cost millions lead to the same intelligence layer.
01
Artist Foundations & Estates
Catalogue, provenance, rights, collectors, and a small staff. Our first and deepest work.
02
Museums
Collections, exhibitions, donors, boards, and infrastructure that has outgrown its documentation.
03
Archives & Libraries
Hundreds of thousands of files and two people able to research them.
04research
Universities & Research Collections
Method, evaluation, and open standards, built together.
PRINCIPLES
Accountable by design.
High-consequence decisions stay human. Every conclusion shows its sources and contradictions. Institutional and collector data remain with their owners. No partner becomes irreplaceable.
HOW IT WORKS
Evidence-native research, verification, and review.
Semantic search finds what an object depicts. We go further: the system connects artworks, documents, people, exhibitions, owners, and publications—including the records that disagree—and shows a scholar why it believes what it believes.
01
Evidence Graph.
Claims remain linked to supporting evidence, contradictory evidence, confidence, provenance, and review status.
02
Adversarial Verification.
A dedicated agent searches for contrary evidence, alternative readings, and source inconsistencies.
03
Human Review.
High-consequence conclusions route to experts with sources, contradictions, confidence, and reasoning visible.
RESEARCH IN PRACTICE
The catalogue is the hard mode.
If a system can reason across provenance, archives, conflicting evidence, private collectors, and scholarly governance, donor workflows and board packets are comparatively straightforward.
EXPLORATORY / RESEARCH PILOT
Tom of Finland Foundation
Catalogue & Digital Archive Initiative
Testing whether multimodal AI, evidence graphs, archival research agents, and structured human review can reduce clerical research burden while the Foundation’s scholarly authority remains final.
Follow the researchSTART HERE
Ready to learn what your institution already knows?
Start with a diagnostic, or talk with us.