Privacy policy.

Product data

Your operational records, pipeline, documents, notes and drafts are never used to train shared models. Data is tenant-isolated, resident in the United States, and governed by single sign-on, role-based access control and audit logging. Model inference runs against providers under agreements that prohibit training on submitted content.

Every connection to a system you run starts read-only. Nothing is written to a system of record without a person on your team approving it, and every write carries a per-field record of who approved it, when, and from which documents.

This site

The public site uses HubSpot for analytics and for booking demos. When you book a demo, the information you provide goes to our HubSpot portal so we can prepare for and run the call. We do not sell your information, and we do not enrich it with third-party data brokers.

To have a HubSpot record removed, write to the address below.

The public record

The public data behind the product is assembled from public sources: federal contracting and spending systems, state procurement portals and transportation programs, and municipal budgets, capital plans and meeting records. Where those documents name individuals acting in a public capacity, such as a contracting officer or an engineer of record, we retain the name as published, alongside a citation to the document.

Retention and deletion

Customer data is retained for the life of the subscription and deleted on request or at termination, subject to any legal hold. Disconnect a system and its data is removed from your model. Public-record data is retained indefinitely, because the value of the record is its history.

Contact

Questions, or a request regarding your own data: hello@capveon.ai.