General Data

Companies that want to buy proprietary datasets for AI training or analysis, and companies that hold data someone else wants, currently have no shared system for sourcing, vetting, pricing and delivering that data safely. Buyers need to confirm a supplier's samples are real, free of personal data, and that the supplier actually holds the rights being sold, while data owners want to license access without handing over their full archive. Today this likely happens through ad hoc emails, manual legal review and one-off contracts, with no single auditable record of who approved what. That leaves both sides exposed on rights, privacy compliance (GDPR, US privacy law, IP) and deal tracking.

General Data provides a pipeline that moves a data transaction through five tracked stages: sourcing via a single intake link, sample review, rights diligence tied to documents and reviewers, contract approval, and controlled delivery. Suppliers respond to a buyer's link with a sample file and no account needed, and can follow deal status through a private link. Every sample is scanned for personal data, which suppliers remove or pseudonymize before any lawful basis is documented. Once all parties sign one dataset version and one license, data moves directly from supplier to buyer, or buyer code runs on the data in place, via file transfer, clean room, or compute-to-data, without General Data storing the full dataset. In-house counsel and data protection officers review each deal for GDPR, the EU AI Act, US privacy law, IP and copyright, and licensing.