Privacy & HIPAA

Your customers' privacy stays in your office. Period.

We de-identify your data on-site using the HIPAA Safe Harbor method, so what gets licensed is no longer personal or protected health information.

The on-site process

Only de-identified data ever leaves your office.

Your systems feed an offline Clean Source device at your location. Identifiers are removed and checked, you approve the result, and only the clean dataset leaves. Raw records, and the drive they were copied to, stay on the premises.

Only de-identified data ever leaves your office Inside your office, data flows from your systems to an offline on-site device that strips identifiers, then you review and approve. Only the de-identified dataset crosses the office line to the AI buyer. Your office: raw data stays inside this line Outside your office Your systems CRM, phones, email, files, scheduling and billing On-site offline device Encrypted, no internet; strips all 18 identifiers You review and approve Spot-check the clean output before anything is released Raw records and the working drive never leave the building. De-identified dataset No names, contact details or IDs; only what you license AI buyer Receives clean data only, then verifies it independently
What is HIPAA Safe Harbor?

A federal checklist, applied to every client.

HIPAA (45 CFR §164.514(b)(2)) gives two ways to de-identify health data. Safe Harbor is the checklist method: remove 18 specific identifiers and have no actual knowledge that what remains could identify someone. Data that meets it is no longer protected health information.

It's the strictest common standard, so we apply it to every business we work with, in healthcare or not.

A small offline processing device on an office desk

The 18 identifiers we remove

  1. Names
  2. Geographic detail smaller than a state
  3. All dates except the year
  4. Phone numbers
  5. Fax numbers
  6. Email addresses
  7. Social Security numbers
  8. Medical record numbers
  9. Health plan beneficiary numbers
  10. Account numbers
  11. License and certificate numbers
  12. Vehicle identifiers and plates
  13. Device identifiers and serial numbers
  14. Web URLs
  15. IP addresses
  16. Biometric identifiers
  17. Full-face photos and comparable images
  18. Any other unique identifying number or code
Our safeguards

Built so there's nothing to leak.

Offline device

No internet connection during processing.

Business Associate Agreement

Signed before we touch anything, for HIPAA-covered clients.

Encrypted, then wiped

Working drives are encrypted and wiped after each engagement.

You approve the output

Review a sample of the clean data before anything is released.

Written summary

A de-identification summary for your records.

Images and files too

Scans and images can carry names and dates in metadata or in the image itself. We scrub both.

Two layers of protection

We clean it. Then the buyer checks it again.

Many brokers send data to the AI company and let the buyer handle de-identification. With us, your data is de-identified inside your office first, and every buyer then runs its own independent privacy review before accepting it.

1
Clean Source Data, on-siteAll 18 identifiers removed on an offline device in your office. You approve a sample before release.
2
The AI company, independentlyThe buyer's own team analyzes the package to confirm nothing sensitive remains. Anything flagged comes back to us to fix.
Not in healthcare?

The same protection for every business.

The same on-site process removes customer names, contact details, account numbers and other personal information from CRM, call, email and dispatch data, in line with state privacy laws such as the CCPA.

Typical broker

A buyer's engineer is given access to your systems to anonymize and extract the data.

Clean Source Data

Buyers never touch your systems or your raw data. Ever.

Want to see how it would work in your office?

Book a short privacy walkthrough. We'll answer every question before anything happens.