← All companies

Centaur AI

Boston company that labels medical and other data through scored contests among human annotators, specialists and AI models, sells model evaluation services, and licenses finished labeled datasets, such as chest X-ray and pathology sets, to AI developers.

Also known as Centaur.ai, Centaur Labs, Centaur Diagnostics, Inc., DiagnosUs

Company website ↗

Source profile checked Oct 29, 2026 · snapshot exported Nov 4, 2026. Importing has not reverified these facts. Original sourced profile ↗

What they do

Description

Checked Oct 16, 2026
Boston company that labels medical and other data through scored contests among human annotators, specialists and AI models, sells model evaluation services, and licenses finished labeled datasets, such as chest X-ray and pathology sets, to AI developers.
See sources (4)

Data handling and security

SOC 2 statement

Checked Oct 16, 2026
Vendor-stated Type II

Company statement; this does not establish an independently audited report.

See sources (1)

Centaur AI Achieves SOC 2 Type II Certification — Centaur.ai ↗

Centaur.ai is proud to announce the successful completion of its SOC 2 Type II audit, further demonstrating the company’s commitment to data security, privacy, and operational excellence.

Original fact status: listed. Check date is separate from import time.

Trust center

Checked Oct 16, 2026
Not found publicly
See sources (0)

No public source was supplied for this value.

Original fact status: not_listed. Check date is separate from import time.

Where it runs

Checked Oct 16, 2026
Not found publicly
See sources (0)

No public source was supplied for this value.

Original fact status: not_listed. Check date is separate from import time.

Data sources

Checked Oct 16, 2026
Built from customer data, Expert or contractor-created, Public / open data
See sources (3)

Arena Datasets — Centaur.ai ↗

Centaur.ai - Expert Data Annotation for Healthcare AI and more ↗

3 New Datasets Added to Centaur’s Arena — Centaur.ai ↗

Built from customer data
Your cases become a private contest. We recruit and gate the right specialists, enter the models you care about, and hand back aggregation labels.
Evidence source ↗
Expert or contractor-created
Select from 100,000+ subject matter experts and/or AI models and have them annotate independently.
Evidence source ↗
Public / open data
1,733 human annotators reviewed 5,856 microscope images from PatchCamelyon, checking whether the marked area contained cancer cells and contributing 209,828 answers.
Evidence source ↗

Original fact status: listed. Check date is separate from import time.

Exclusivity

Checked Oct 16, 2026
Not found publicly
See sources (0)

No public source was supplied for this value.

Original fact status: not_listed. Check date is separate from import time.

Company information