ScrubPHI Guard — Founding Pilot
Your staff already use AI tools.
Guard the PHI. Don't ban the tool.
Care coordinators, schedulers, and clinicians already paste real people's details into public AI chat tools — names, dates of birth, diagnoses, visit notes. A ban sounds clean and doesn't survive the week. ScrubPHI Guard sits at the moment of paste-and-send in the browser, detects protected health information, and holds it before it leaves your organization — so your people keep the productivity and you keep the PHI.
Request the free 60-day pilotFor compliance officers at small HIPAA-covered practices and care programs. No server to install, no data collected by us.
How it works
A browser extension your IT or compliance lead deploys with a policy template. It runs in Chrome and covers ChatGPT, Claude, Gemini, and Copilot. Everything happens on the staff member's own machine.
01
Detects on-device
Guard watches text headed for AI chat tools and recognizes PHI patterns — names paired with dates, MRNs, phone numbers, addresses, towns, and facility names such as hospitals and nursing homes. Detection runs entirely in the browser. There is no Guard server, and your text never comes to us.
02
Holds before send
When text looks like PHI, Guard holds the paste or send and shows what it found, by type, with a preview of the scrubbed version. Staff insert the scrubbed text — identifiers replaced with markers like [REDACTED:TYPE] — or cancel. In warn mode, staff can choose to keep the original, and that choice is logged.
03
Counts-only audit
Your compliance file gets evidence, not contents. Guard logs counts — how many holds, of what type, when — exportable as CSV. No flagged text is stored in the log. You can show an auditor the guard is working without creating a new pile of PHI to protect.
The founding pilot
We are enrolling a small founding cohort now. Guard is pre-release (v0.3), and the pilot is how we prove it in real programs before general release.
- $0 for 60 days — full extension build for your organization.
- Admin policy template — deployment settings your compliance lead controls.
- Setup call included — we set block or warn mode, the covered-site list, and the policy template with you.
- After the pilot: simple flat per-organization pricing, agreed with you before the pilot ends. No per-seat counting.
Stated plainly
Detection is best-effort. No filter catches everything, and Guard is a guardrail, not a guarantee — flagged text still deserves a human review before use, and your staff still need a clear AI-use policy behind the tool.
Guard works on text. Images, files, and screenshots are not scanned; a notice appears when a file is pasted, and your policy can block file pastes outright.
Guard reduces the chance that PHI reaches a public AI tool. It does not make a non-compliant tool compliant, and it is not legal advice. What it gives a compliance officer is a working middle option between "ban it and be ignored" and "allow it and hope."
Who builds it
ScrubPHI Guard is built by Hammertime Consulting LLC, a veteran-owned Pennsylvania company, as part of the SI App Factory. It grew out of the same conviction as our other tools: small, well-made software that solves a real problem for real people, without noise. Questions go to a person, not a ticket queue: eric@hammertimeconsulting.co.