AI governance · independent clinics · RN‑led
Your staff are already using AI. Nobody’s watching the vitals.
I find where AI is already being used in your practice, put the agreements, policy and patient disclosures in place, build one workflow your team actually wants, and train them on it. Then I keep watch.
- Fixed prices
- You keep everything
- On-site across Central Texas
Illustration: an example clinic before governance.
01What’s already happening
The AI is already inside your clinic. The question is where the patient information goes.
Staff use whatever saves them time: a personal ChatGPT login, a free transcription app, a browser extension that summarizes faxes. None of those come with a business associate agreement, and nobody reviews what comes back. Flip the switch to see the same clinic with governance in place.
- Front desk: summarizing voicemails and drafting replies to patient messages.
- Exam rooms: rewording discharge instructions and after-visit summaries.
- Billing: drafting appeal letters from denial notices.
- Referrals: chasing outside records and summarizing faxes.
- Office: writing policies, schedules and staff emails.
- Approved tools only. Covered by agreements that include the features your staff actually use.
- A person reviews what matters. Anything that reaches a patient or a payer gets checked first.
- Everyone knows the rules. What goes in, what stays out, and who to ask at 4:45 on a Friday.
02Why now
In Texas, telling patients is now the law.
Two state laws changed what practices owe patients when AI touches their care. If your intake forms, consent language and staff training haven’t changed since 2025, start here.
-
Texas SB 1188
If AI helps with a diagnosis, the practitioner must review the AI-generated record and tell the patient that AI was used.
-
Texas HB 149 (TRAIGA)
Providers must tell patients, in plain language, when AI is used in relation to their care, no later than when that care is first provided (emergencies excepted).
-
81% of physicians
report using AI, in a survey of 1,692 physicians.
AMA -
8% of nurses
say their employer has communicated a clear AI strategy for their role.
Incredible Health -
Your clinic
Is your AI use written down, disclosed to patients and reviewed by a person?
Take your clinic’s vitals
General information, not legal advice. Confirm what applies to your practice with your healthcare attorney. Law summaries: Texas Medical Association, TMLT. Surveys are separate and self-reported.
03Vitals check
Take your clinic’s vitals. Seven questions, two minutes.
Answer for your practice as it runs today. The monitor responds as you go, and you get an acuity level at the end.
Question 1 of 7
04What governed looks like
Governed AI knows its sources, its dates and its limits.
Every workflow I build is tested for these three behaviors before your staff rely on it. Here they are in a fictional session.
Fictional session at the referrals desk. The assistant answers from the Outside Records Follow-up procedure and cites its source and owner.
05The care plan
Assess. Treat. Monitor.
The same way you’d care for a patient: find out what’s going on, fix what matters most, then keep watching. Fixed prices, written scope, and you keep everything.
3 spots at the founding rate for Setup, in exchange for a case study once your workflow is live.
-
01 · Assess1–2 weeks
AI Readiness Review
$950
Credited toward Setup if you continue within 60 days.
- Inventory of every AI tool in use, including personal accounts
- BAA and feature check for up to five tools that could see patient information
- A one-page AI policy your staff can follow
- Draft patient AI disclosure for intake and consent, for your attorney to review
- A prioritized plan: what to fix first, and what to build
-
02 · Treatabout 1 month
Clinic AI Setup
$4,500standard $6,000
Founding rate. 50% to start, 50% at handoff.
- Everything in Assess
- Up to four staff interviews, including the workarounds
- A clinic-owned knowledge pack from your approved procedures
- One workflow on an approved platform, tested against realistic cases, outdated documents and questions it should refuse
- Staff training and a controlled launch with a fallback and a named owner
- Copies of everything at handoff
-
03 · Monitormonthly, optional
Ongoing governance
$600/ month
Stop anytime and keep everything. Terms are in the agreement.
- Quarterly governance review
- Updated policy and disclosure drafts when laws or vendor terms change
- Review of new AI tool requests from staff
- AI training for every new hire
- Everything in Keep current
- A monthly review of use and reported issues
- One new or improved workflow each month
- Staff refreshers when a workflow changes
One location per engagement. If you need a new software license, the cost and the reason are in writing before anything is bought. Extra workflows and integrations are scoped separately.
06Order sets
Start with one workflow your team will actually use.
Setup includes one. Pick the one that costs your staff the most time.
-
Billing
Prior auth & appeal drafts
Drafts appeal letters from the denial notice and the notes you approve. A person reviews every letter before it goes out.
-
Front desk · Clinical
Policy answers for staff
Answers questions from your own procedures, cites the page, and sends anything unclear to the person who owns it.
-
Clinical
Patient message drafts
Drafts portal replies in your clinic’s voice. Anything clinical goes to a clinician, every time.
-
Referrals
Records & referral follow-up
Tracks outstanding records, drafts the follow-ups and flags what’s overdue before a patient notices.
-
Office
New-hire onboarding
Turns your procedures into a guided first week, with quick checks along the way, so training doesn’t depend on who’s free.
-
Front desk · Texas
Patient AI disclosure
Plain-language disclosure for intake, consent and signage, matched to how your clinic actually uses AI.
John Lewis
Registered Nurse · Founder
If found, return to a clinic that wants safe AI.
- License
- Registered Nurse
- Work
- Health data operations
- Cert
- HIMSS: Assessing & Implementing AI and ML in Healthcare
- Base
- Austin, TX
john@clinicalaigovernance.org
Swing the badge · tap to flip
07Who you’ll work with
I’m a nurse who learned to build.
I’m John Lewis, a registered nurse working in health data operations. I build the databases, Python tools, workflows and interactive training that keep a healthcare operations team running, including training for people who had never worked inside the U.S. healthcare system. AI is how I learned to build all of it.
That’s what I bring to your clinic: someone who understands how a front desk, a billing queue and a referral backlog actually work, and who can build the tool that eases them.
I still think like a nurse. Before anything goes live, I ask what happens at 4:45 on a Friday, when the usual reviewer has gone home and someone still needs an answer. Then we build that answer into the instructions, the review process and the training.
- LicenseRegistered Nurse
- CertificateHIMSS · Assessing and Implementing AI and Machine Learning in Healthcare
- ServesSolo and group independent practices
- WhereAustin, TX · on-site across Central Texas, remote everywhere else
- ProfileJohn Lewis, RN on LinkedIn
I take on a few clinics at a time, so the person you meet on the first call is the person who does the work.
What happens on the first call?
We talk through the tools your staff use, the work they want help with and who needs to be involved. It’s a fit conversation. A written scope and price come before any commitment.
Will you sign a business associate agreement?
Yes. If the work involves access to patient information, we sign a BAA before it starts.
Does this make us compliant with Texas AI laws?
No one can promise that on a website, and I won’t. What I put in place is the practical groundwork: patient disclosure language for your attorney to review, a clear point where a person checks AI output, and the inventory, policy and training that keep both working. Whether your practice meets a specific legal requirement is a question for your healthcare attorney, and I’m happy to work alongside them.
Do we need to buy new software?
Usually not at first. The review starts with what you already pay for. If a new license is needed, the cost and the reason are in writing before anything is bought.
What if we’re not ready to launch?
You still get the findings, the policy, the knowledge pack and a clear record of what’s blocking launch. If an agreement is missing or a control is untested, the workflow waits, and you get that in writing.
Do we have to keep the monthly service?
No. Monitor is optional, and you keep your policy, knowledge pack, instructions and training materials either way.
Is this legal advice or a compliance certification?
No. It’s operational review and implementation support. Your practice keeps approval authority, legal interpretation goes to your attorney, and clinical decisions stay with your clinicians.
Are you local?
I’m based in Austin. Interviews and training can happen on-site across Central Texas, and everything works remotely anywhere else.
09Next step
This is what monitored looks like.
Bring the repeated question, the awkward handoff or the tool nobody approved. The first call is 30 minutes, about your clinic, with no commitment.