AI for medical practices, with clinicians in charge

What we would build with a practice

These are proposed services. We have not yet delivered them for a healthcare client. Each one starts with onboarding, where we find out whether it is worth building for your practice at all.

How a practice starts

Every client follows the same process. After a free sales call, you pay a fixed $10,000 for onboarding. We write a questionnaire for your practice. On the onboarding call, the front desk, medical assistants, billing staff, and clinicians each walk us through a normal day. We map where the hours go, confirm what your EHR vendor and other systems allow, and deliver an ontology within 72 hours for one workflow worth building first. We then present recommendations. Certified engineers build inside the systems you already run, under a separate contract billed monthly, and we measure the result against the baseline we took. More on how we work.

Clinicians stay in charge

AI reads, sorts, drafts, and flags what is missing. Software applies the rules and does the arithmetic. A licensed clinician makes every clinical decision and signs every note, and staff approve anything that goes to a patient or a payer. Nothing we build diagnoses, prescribes, places orders, or sends clinical advice to a patient. It is the same split that made our construction tools get used: AI never supplied the prices there, and it will not supply the diagnosis here.

Physicians want this too. In the AMA's 2026 survey, 85% wanted to be consulted on or directly involved in AI adoption decisions, 88% called safety and efficacy validation critical for broader adoption, and 86% said the same of data privacy. Clinicians help choose what gets built, review what it produces, and can switch it off. If a tool records visits, we recommend telling the patient at the start and having your counsel confirm what consent your state requires.

Source: AMA, 2026 physician survey on augmented intelligence

Patient data and access

BAAs with vendors that handle ePHI

HHS guidance says a cloud provider that creates, receives, maintains, or transmits electronic protected health information (ePHI) for a practice is a business associate, even if it stores only encrypted ePHI without the key, and that using one without a business associate agreement violates the HIPAA Rules. One that handles only de-identified information is not. So before ePHI enters any tool, a practice should have a BAA with each vendor in the data path that qualifies. Whether a project needs Mandala to access ePHI at all is assessed per project, and we aim to design so it does not. If it does, your privacy officer and counsel decide what agreement is required first.

Source: HHS, Guidance on HIPAA and cloud computing

Access agreed in writing

HHS notes that security responsibilities such as access controls can be split between a practice and a cloud provider, with each confirming in writing how it addresses the Security Rule. We map who can see what, where data is stored, and for how long, so your practice and its vendors can put that in writing before anything is built.

Source: HHS, Guidance on HIPAA and cloud computing

Built without patient data where the work allows

Onboarding starts with walkthroughs of how the work happens. Where the work allows, we design and test on synthetic or de-identified examples and ask for the least access the work needs.

Please do not send patient information through this site's chat, contact form, or booking form.

What the research shows so far

Most physicians report some professional use of AI

In the AMA's 2026 survey, which analyzed responses from 1,692 physicians, 81% said they use AI professionally, more than double the rate when the AMA first asked in 2023. That counts any professional use; the most common was summarizing research and standards of care (39%). Documentation and messaging uses were each less common: discharge instructions, care plans, or progress notes (30%); billing codes, charts, or visit notes (28%); and draft portal replies (19%). Medscape noted that the AMA did not explain how the survey population was selected.

Source: AMA, 2026 physician survey on augmented intelligence; Medscape coverage of the AMA survey

Ambient scribes were associated with modest time reductions

A JAMA study of 8,581 ambulatory clinicians at five academic health systems, 1,809 of whom adopted an ambient AI scribe, found adoption was associated with 16 fewer minutes of documentation and 13.4 fewer minutes in the EHR per eight scheduled patient hours. After-hours EHR time did not change significantly. Mass General Brigham, one of the sites, called the reductions modest: about 10% of documentation time. Larger changes were seen in groups including primary care and clinicians who used a scribe in at least half of their visits, and only 32% of users used it that often. These are associations at large academic systems. None of them forecasts results for any one practice.

Source: JAMA, 2026 multisite study of ambient AI scribes; Mass General Brigham press release

Questions practices ask first

Have you done this for a medical practice before?

Not yet. Our delivered work is in construction, where we built a custom Procore application inside CJM Construction in South Florida. We are now meeting with physician practices and clinics. We will tell you plainly what has been proven and what has not, and every engineer in the network follows the same process.

Is your work HIPAA compliant?

There is no official HIPAA certification to show you: HHS's Office for Civil Rights says it does not endorse, certify, or recommend specific technology or products, so a "HIPAA compliant" label settles nothing on its own. Compliance depends on the agreements in place, the safeguards, and how tools are configured and used. We are engineers and cannot act as your privacy officer or counsel, and we make no legal or compliance guarantees. We give them the data-flow map and vendor list they need to make the call.

Do we have to replace our EHR?

No. We build around the systems you already run. What we can connect to depends on what your EHR vendor allows, whether that is an API, scheduled exports, or reports. We confirm that during onboarding before we commit to building anything.

How much time will this save us?

We will not promise a number. In a multisite JAMA study of ambient scribes, average changes were modest and varied by specialty and by how often clinicians used the tool. We measure a baseline during onboarding and report against it.

Can our staff paste patient information into a chatbot?

Not into one your practice has no BAA with. HHS guidance treats a cloud service that creates, receives, maintains, or transmits ePHI for a practice as a business associate that needs a BAA. We can help your compliance team write the policy and choose an approved tool so staff have somewhere appropriate to work.