How does AI know what to write in the note before the appointment happens?

It pulls the scheduled procedure, the patient's history, and prior notes from your PMS the night before to build a draft, then updates that draft during the visit based on what's actually said and done in the room. The draft is a starting point built from real data, not a generic template, and it's edited before signing to match the actual visit.

Do I still need to review every AI-generated note before signing?

Yes, and the workflow is built around that review — every note requires a dentist's read-through and sign-off, typically 30 to 60 seconds per patient, before it's finalized in the chart. The system drafts the note; the licensed provider remains responsible for its accuracy and signature.

Will AI-drafted notes hold up if a payer audits a claim?

AI-drafted notes are built against procedure-specific completeness checklists so the narrative includes the documentation elements payers typically request, such as medical necessity language for periodontal claims or tooth-specific detail for restorative work. That said, the dentist's review before signing is what confirms the note is accurate and complete for that specific patient, which is what actually protects the claim.

What happens if a patient is seen but the note never gets finished?

The system compares the day's completed schedule against signed notes at a set point each day and flags any seen patient without a completed note, so the gap gets caught the same afternoon instead of weeks later during a claim denial or audit. That flag goes to whoever manages the clinical schedule, so the note gets finished while the visit is still fresh.

Can I still dictate parts of the note manually?

Yes — dictation stays available for adding specific findings, informed-consent conversations, or follow-up notes directly into the draft. It's meant to layer on top of the automatically drafted note for exceptions, not replace the drafting process for the whole visit.

Does this replace or integrate with my current practice management system?

It integrates with and writes back into your existing PMS rather than replacing it, so your team keeps using the system they already know for scheduling, billing, and the permanent chart. Notes, along with insurance verification and intake data, sync into the record you already maintain.

How Much Time Do AI Clinical Notes Save Per Op Day?

On a typical 20-patient operatory day, AI-drafted clinical notes save a dentist roughly 45-60 minutes of charting time, because the SOAP note for each patient is already written — pulled from the schedule, the treatment plan, and the prior note — before you walk into the room. You read it, correct anything that doesn't match what actually happened, and sign. What used to be 4-6 minutes of typing per patient becomes 30-60 seconds of review.

We built AI clinical notes for our own offices because charting was the single biggest reason we were still at the office at 6:45 on a Tuesday. This post is the math and the mechanics: what gets written automatically, what still needs a human decision, and how a practice catches the notes that would otherwise slip through unsigned.

How much time do AI clinical notes save per op day?

Start with what a note actually costs you today. A dentist typing a SOAP note from scratch — chief complaint, exam findings, diagnosis, procedure detail, materials, anesthesia, post-op instructions — takes 3 to 6 minutes if the day is calm and closer to 8 if you're trying to remember what you did three patients ago because you got behind. Multiply that across a full schedule and it's not a rounding error, it's an hour or more of unpaid clinical labor stacked onto the end of your day.

The math

Take a realistic single-provider day: 18 patients, average 5 minutes per note if typed manually.

18 notes x 5 minutes = 90 minutes of manual charting per day
With AI-drafted notes, review-and-sign averages 45 seconds per note: 18 x 0.75 minutes = 13.5 minutes
Net savings: roughly 76 minutes per clinical day
Over a 20-day clinical month, that's about 25 hours — more than three full clinical days back

For a two-provider practice, double it: 50 hours a month that were previously spent finishing notes after the last patient left, transcribing handwritten shorthand, or catching up on a Saturday. That time either goes home with you or goes back into the schedule as production capacity — an extra half-day of patients a month without adding an operatory or a chair.

What's already written before you sit down

The system pulls the day's schedule from your PMS the night before and pre-populates a draft note for every scheduled procedure: patient history, the planned treatment code, prior notes for context (was this a re-prep, a follow-up on a diagnosis, a continuation of a quad of SRP), and known allergies or medical alerts. For a scheduled crown prep, the draft already has the tooth number, the planned material, and standard anesthesia language pulled from your own historical patterns for that procedure type — not a generic template, but one built from how you and your associates actually chart.

During the appointment, the assistant listens (with consent, per your state's two-party or one-party recording rules) and updates the draft with what was actually said and done — findings, deviations from the plan, patient responses, anything that changes the diagnosis or treatment rendered. By the time you're washing your hands, the note reflects the visit, not just the schedule.

Chart-ready, not just typed

"Chart-ready" means the note is formatted the way your PMS expects it and the way an auditor or insurance reviewer would want to see it — narrative SOAP structure, procedure codes cross-referenced to the narrative, and the specific documentation elements payers ask for when they request records (medical necessity language for periodontal procedures, tooth-specific detail for restorative work, radiographic findings tied to the diagnosis). That last piece matters more than it sounds: a note that's technically complete but missing the two sentences a payer wants is functionally the same as no note when a claim gets audited.

Note completeness and compliance

Most incomplete notes aren't wrong, they're thin — a procedure code with one line of narrative and nothing else, written at 5:40 pm by someone who wants to go home. That's the note that gets flagged in a payer audit or a state board review two years later when nobody remembers the visit. AI-drafted notes are built against a completeness checklist specific to procedure type, so a note for periodontal surgery includes the elements a periodontal note needs, and a note for an emergency exam includes the elements that support the diagnosis and justify the treatment billed.

This matters even more when it's paired with automatic insurance verification — if the estimate you gave the patient assumed a certain diagnosis code, the clinical note needs to support that code, or the claim comes back downgraded or denied. Complete, procedure-matched documentation is one of the more direct ways clinical notes affect collections, not just compliance.

Missing-note detection

Every PMS has patients who were seen, treated, and checked out with no note ever finished — usually because the day got busy and it fell off the list. The system checks the day's completed schedule against signed notes at a set time each afternoon and flags any patient who was seen but doesn't have a completed, signed note yet. Instead of finding this three weeks later when a claim gets denied for lack of documentation, the front desk or clinical lead sees it before the day closes, while the visit is still fresh enough for the dentist to write an accurate note rather than reconstruct one from memory.

Where dictation still fits

Ambient drafting from the schedule and the conversation in the room handles most of the note. But there are moments where you want to say something directly and have it go straight into the record — an unusual finding, a specific informed-consent conversation, a note to yourself about following up with the patient's physician. Dictation remains available for exactly that: short, deliberate additions layered onto the draft, not a replacement for typing an entire note from a blank page. The two modes work together — draft from the schedule and conversation, dictate the exceptions.

Where the saved time actually goes

The hour or so recovered per day shows up in a few places depending on the practice. Some dentists use it to see one more patient a day without extending hours. Some use it to actually eat lunch. A few have used the freed-up chair time to add virtual consultations at the start or end of the day — quick screening visits that used to get squeezed out because there was no time between patients to also finish notes. None of those uses require hiring anyone; they require the notes not eating the gap between patients in the first place.

It runs on top of your existing system

None of this requires replacing your PMS. AI clinical notes, along with the rest of the AI front office — receptionist, intake forms, insurance verification, scheduling — sit on top of the system you already use and write back into it, so your team isn't learning a second platform or double-entering data. The chart lives where it's always lived; it just arrives mostly written.

If you want to see what a drafted note looks like against your own procedure mix, schedule a demo with a few of your actual codes and a representative day. Pricing and what's included at each tier is on the pricing page.

The bottom line

Charting is unavoidable — it's the record of care and the backbone of every claim you submit. What's avoidable is spending 90 minutes a day typing narratives from a blank screen when the schedule, the treatment plan, and the conversation in the room already contain everything the note needs to say. Pre-drafted, chart-ready notes don't remove the dentist from documentation; they remove the typing.