Data Sharing — How Vulpa Uses Your Information
Last updated: May 1, 2026
The short version
If you opt in to “Help improve Vulpa,” we use anonymized patterns from your project documents to make the AI smarter for every contractor. We never share your specific documents, project names, addresses, or any identifying details with other users.
If you don’t opt in, your data is used only to power your own projects. Nothing leaves your account.
You can change your mind anytime on your Profile page.
What gets shared (when opted in)
- Anonymized correction patterns — e.g., “30% of King County addition permits get flagged for missing structural calcs” — with no project, contractor, or address attached.
- Aggregated language patterns — common phrasings that triggered reviewer corrections, stripped of identifying context.
What is NEVER shared
- Your project documents (PDFs, drawings, plans)
- Project names, addresses, or parcel numbers
- Your name, company name, or contact info
- Client names or any third-party identifying info
- Pricing, contract terms, or financial details
How removal works
When you turn off data sharing, your data is removed from active systems within 30 days and excluded from future processing. Residual data in backups is retained only for disaster recovery and is deleted per standard retention cycles.
We cannot remove data from previously generated aggregates or from any AI model weights that may have been influenced by the data, and we do not claim otherwise.
Why this matters
Vulpa’s accuracy improves when contractors share patterns. The Seattle SDCI public dataset gets us started — your opted-in contributions sharpen the AI for the specific work you do. Your individual documents are yours alone, always.
Questions? Email support@vulpa.ai.
Third-party processors
Vulpa uses Anthropic’s Claude AI as part of our processing pipeline, including the anonymization step that runs before any data is added to our shared knowledge base. Anthropic acts as a data processor under our Data Processing Addendum. They do not use customer data to train their models, do not reuse data across customers, and do not monetize customer content. The processing relationship is governed by Anthropic’s Commercial Terms of Service and the linked DPA at anthropic.com/legal/data-processing-addendum.