Use case
AI for medical clinics, on hardware the practice owns
HIPAA aware intake, a front end on the EHR you already use, multi physician scheduling, and retrieval over your own treatment protocols. We are not the EHR and we are not a covered entity. We build the layer that makes the EHR usable for patients and staff, and we build it so patient information never leaves the practice to be read by an outside model.
The rules constrain the surface. Staff time is the real bottleneck
Clinics carry a specific kind of pain: patient intake done by hand on a clipboard, front desk phone load with the same five questions, an EHR patient portal patients quietly refuse to use, and referral coordination that still moves by fax.
A new patient fills a clipboard and the front desk retypes it into the EHR, twenty minutes of staff time per intake plus the risk of a typo. "What insurance do you take," "how do I refill," "what is the address," asked daily and answered by a human every time. Patients are told to log into a portal, fail the password reset, and give up, so a friendlier front end changes adoption more than any reminder email does. A specialist needs records from a primary care office and they still move by fax, days lost for no clinical reason.
None of that requires handing patient data to a hosted model to fix. It requires a system that runs on infrastructure the practice already trusts.
Sources: HHS HIPAA Security Rule, HHS HIPAA Breach Notification Rule.
What it does
Five capabilities that respect HIPAA and improve patient experience
HIPAA aware intake
Patients fill intake on a phone before the visit, consents signed, insurance card captured, all encrypted at rest and in transit on infrastructure the practice controls rather than a rented form tool.
A front end on top of the EHR
A clean, modern interface for portal access, messaging, refills, and scheduling. The EHR stays the system of record. We build the experience layer on top of it.
Multi physician scheduling
Scheduling across multiple providers and locations, with authorization status visible up front, which reduces same day cancellations from surprises at check in.
Protocol retrieval for staff
Clinical staff query your private treatment protocols and published guidance in plain language, with a citation every time, and nothing patient identifiable enters the query.
Site chatbot for the public site
Address, hours, accepted insurance, what to bring, refill policy, all cited from your published policy. It does not give medical advice.
Real work, already shipped
Northwest Arkansas Infectious Disease is a specialist practice serving the region. We built the public clinic site as a hand coded property with full clinic and physician schema, provider bios, and accepted insurance clearly published, alongside the search visibility work that helps new patients and referring physicians find the practice. The patient portal and EHR stayed on the practice's own EHR system throughout, and the public site never stored patient health information.
That separation, public facing content on one side and patient data on the practice's own systems on the other, is the same boundary a private federation formalizes and extends into the intake and scheduling layer.
Medical FAQ
Buyer questions, medical specific
Is That AI Guy HIPAA compliant?
That AI Guy is not a covered entity. HIPAA obligations run through the covered entity, your clinic, and any business associate that touches patient data. A private federation removes a class of that exposure by keeping intake, scheduling, and protocol data on hardware the clinic owns, so the data is never shipped to an outside model to be read. Where any outside service is still involved, a Business Associate Agreement is the real instrument, not a marketing claim.
Can the AI give medical advice?
No. AI on a clinic site is constrained to operational questions: address, hours, insurance accepted, refill policy, what to bring to a visit. Medical advice comes from the clinician, always.
Will you replace our EHR?
No. EHR replacement is a multi year, multi vendor decision that belongs to the practice. We build the experience layer on top of the EHR you already have.
What about state medical board advertising rules?
Site copy follows your state medical board rules. Specialist titles, board certifications, and outcome claims are reviewed against board guidelines before anything publishes.
How is insurance card capture handled securely?
Photos go to encrypted storage on infrastructure the practice controls, reachable only by authorized staff through an authenticated session. Images do not pass through a public model API on the way in.
Other industries
Verticals with the same shape of problem
Begin
Should your practice own its AI?
The readiness assessment answers exactly that, for your intake, your EHR, and your staff workflow, and ends with a real plan.