HANS vs Generic AI Scribes

What's wrong with using generic AI scribes for functional medicine?

The same thing that's wrong with using a screwdriver as a hammer. They're both tools, but they're built for completely different jobs.

Every major AI scribe on the market — Nuance DAX, Suki, Abridge, and the rest — was designed for conventional medicine. Their core use case: document a 15-minute office visit, transcribe the conversation accurately, produce a standard SOAP note. These are genuinely useful tools for primary care, urgent care, and specialties that work within that framework.

Functional medicine isn't that. Your visits are 60-90 minutes. Your notes require synthesizing multiple lab panels, tracking longitudinal trends, and connecting findings across body systems in ways that standard documentation doesn't support. The "subjective" section of a functional medicine note could easily be 500 words covering sleep, diet, stress, environmental exposures, and symptom patterns. The "assessment" is rarely a single diagnosis — it's a systems-biology synthesis.

Generic AI scribes can't handle this. Not won't — can't. It's not a defect in the technology; it's that the technology was never trained on functional medicine data.

What happens when you use generic AI scribes with functional medicine labs?

Specific problems. Not hypothetical ones — we hear about these from practitioners every week:

• DUTCH hormone panels: Generic AI consistently misinterprets cortisol metabolites. It confuses free cortisol with total cortisol metabolites, mislabels elevated 11ß-HSD activity as "normal stress response," and entirely misses the clinical significance of the cortisol/DHEA ratio in the context of HPA axis dysregulation.

• Organic Acids Test (OAT): When a patient has elevated quinolinic acid with low serotonin metabolites, that's a critical finding pointing toward neuroinflammation and the need for tryptophan support versus direct serotonin support. Generic AI either ignores it (produces a note that says "OAT results reviewed, appears normal") or describes it in generic terms that lose clinical meaning.

• GI-MAP: Generic AI struggles with the clinical significance of opportunistic pathogens versus commensal flora overgrowth. It doesn't know that SIgA elevation in the context of zonulin positivity points toward different interventions than SIgA elevation alone.

• BaleDoneen protocol documentation: You can't document cardiovascular risk the same way in functional medicine. You're tracking Lp(a), ApoB/ApoA1 ratios, LDL particle number, and inflammatory markers — then connecting them to dietary recommendations, targeted nutraceuticals, and lifestyle interventions. Generic AI produces notes that sound like they came from a conventional cardiologist.

• IFM-style case presentations: When you're documenting a case using functional medicine methodology — timeline, matrix, modifiable lifestyle factors — generic AI flattens it into generic problem lists that lose the narrative thread that makes FM documentation clinically useful.

How is HANS different?

We built HANS specifically for functional medicine. Our models were trained on actual functional medicine charts, not adapted from conventional medical documentation.

Here's what that means in practice:

Capability	Generic AI Scribe	HANS
DUTCH hormone interpretation	Confuses metabolites, misses ratios	Understands cortisol/DHEA relationships, HPA axis patterns
OAT analysis	Generic "results reviewed"	Flags specific metabolic patterns (quinolinic acid, kynurenine pathway)
GI-MAP interpretation	Misses opportunistic vs. commensal distinction	Understands clinical significance of pathogen combinations
Longitudinal tracking	Treats each note as standalone	Surfaces trends across visits and labs
FM-native documentation	Forces SOAP format	Uses functional medicine narrative structure
Protocol tracking	None	Tracks supplements, lifestyle changes over time

HANS saves 30 minutes per note and reduces patient prep from 2 hours to 10 minutes. That's not theoretical — we measured it with practitioners in real practices.

What if I only need basic transcription?

If you only need someone to accurately transcribe your dictation, any AI scribe will do. But that's not what functional medicine requires.

Your notes need to do more than accurately record what you said. They need to capture the clinical significance of what you found and connect it to what you're planning. They need to surface relevant history without forcing you to re-enter everything. They need to track treatment progress over time.

A generic AI scribe will give you an accurate transcript. HANS gives you documentation that helps you practice better medicine.

Do I lose anything by switching from a generic AI scribe?

You lose the frustration of notes that don't reflect your actual clinical thinking. You lose the time spent editing generic output to make it clinically useful. You lose the patients you didn't see because you were doing chart prep for three hours after your last appointment.

What you gain: evenings back. More capacity for patients. Notes that actually help you track cases over time.

Who is this for?

If you're a functional medicine practitioner spending more time on documentation than patient care, you've already identified the problem. Generic AI scribes were never the solution — they're tools built for a different type of practice.

HANS is built for FM practitioners who want to practice medicine, not documentation.