Protege
Protege is an AI data platform that aggregates proprietary, real-world datasets from hundreds of partners and prepares them for AI model training, fine-tuning, and evaluation. The company curates and de-identifies data across healthcare, media, audio, and motion capture, then delivers it to AI builders under rights-protected licensing agreements. Its catalog includes billions of clinical notes, hundreds of millions of medical images, and over 500,000 hours of audio content.
- Artificial Intelligence
- Data & Analytics
- Developer Tools
- Software Only
Funding
Raised to date
$65MRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
across 3 rounds
+4$30M · Series A Extension
Announced Jan 2026 · from article date
Disclosed cumulative: $65M
Founders
Travis May
Co-Founder & Chairman
Protege • Co-Founder & Chairman
Product
Problem
AI development is increasingly bottlenecked by the scarcity of high-quality, non-public data needed for training, fine-tuning, and evaluation. Real-world data is highly fragmented across industries, and neither data holders nor AI builders are equipped to operationalize it at scale, while publicly available datasets often suffer from contamination, lack of multimodal depth, and poor reflection of real-world workflows.
Solution
Protege serves as the connective tissue between data providers and AI model builders, aggregating proprietary datasets from hundreds of partners and preparing them for use across the entire AI lifecycle. The company curates, de-identifies, and quality-checks data to create AI-ready datasets tailored to specific model-building needs, covering pre-training, post-training, fine-tuning, and evaluation. Protege's Data Lab provides domain-specific expertise to solve complex AI development challenges, while its secure delivery infrastructure ensures rights and privacy protections. Data providers unlock new revenue streams through standard licensing agreements, and AI builders gain access to hard-to-find, high-quality data that reflects real-world conditions.
Target Audience
Primary customers are AI model builders, including vertical AI companies, large foundation model developers, and in-house AI teams, as well as ecosystem enablers such as annotation and labeling companies and benchmark builders.
Features
- Aggregated access to proprietary datasets across healthcare, media, audio, motion capture, energy, manufacturing, real estate, finance, and agriculture
- Catalog includes over 300,000 hours of video, 500,000 hours of audio, billions of clinical notes, and hundreds of millions of medical images
- Custom data curation, de-identification, and quality checks to ensure datasets are AI-ready and rights-protected
- Evaluation and benchmark-specific datasets designed to be uncontaminated, multimodal, and reflective of real patient journeys
- DataLab provides expert research and domain-specific data knowledge to solve unique AI development challenges
- Revenue share payouts to data partners with each use of their data