Type a question the way you'd say it out loud ("what was production last month?") and get an answer drawn from your live practice data. It's an AI chat built for a dental office, powered by Anthropic's Claude, and the figures it shows come from the same metric layer as your dashboards.
No report builder, no filters, no spreadsheet. These are ordinary questions with exact answers behind them — and any answer downloads as a one-page PDF for the morning huddle, the partners, or your accountant.
"What was our production last month?"
Net production for any date range, using the same definition as your end-of-day report. Not a sum the model did in its head.
"How's our collection rate this quarter?"
Collections against production minus adjustments, plus a breakdown of what came in by payment type.
"How many new patients did we see in March?"
Counted on the D0150 new-patient rule, so it matches what your dashboard and your accountant already show.
"How much diagnosed treatment is sitting unscheduled?"
The dollar value of treatment that was diagnosed and never booked, trailing twelve months, with reactivation and unused-benefit opportunity alongside it.
"What's our no-show rate, and how booked are the chairs?"
No-show and cancellation rates plus chair utilization, so you can see whether the gaps are patients or scheduling.
"Who has unused insurance benefits before year-end?"
Remaining benefits across your patient base. The list most practices only find in December, when there's no chair time left.
An assistant that invents a number is worse than no assistant. This one is built so it can't.
Every figure is computed by the same canonical metric layer that produces your dashboards and end-of-day reports, then handed to the assistant to explain. What you read can't drift from what your reports say, because it is what your reports say.
If a sync is behind, the answer carries the freshness caveat rather than quietly presenting yesterday's picture as today's. You get the number and the asterisk.
Open Dental, CareStack and Dentrix Ascend are read through one adapter, so the same question works the same way regardless of what runs at the front desk.
It's a chat box in your dashboard that answers questions about your own practice. Instead of building a report, you ask ("how did we do last month?", "what's our case acceptance?") and it queries your practice data and answers in plain language, with the underlying figures shown as cards and tables you can read at a glance. Any answer downloads as a one-page PDF, with every figure recalculated as the report is built.
Anthropic's Claude. It handles the language, understanding the question and explaining the result. It does not calculate your numbers; those come from our metric layer, which is why the answers reconcile with your reports.
The figures on screen are computed before the model ever sees them and rendered from structured results, not from the model's prose. If a metric can't be computed, you're told that instead of being given a confident guess.
It works with Open Dental, CareStack and Dentrix Ascend, and with other supported systems through our integration layer. Questions are phrased the same way whichever one you use.
The assistant runs inside the same HIPAA-compliant platform as the rest of your account, under the same encryption and access controls. Most performance questions are answered from aggregates rather than individual charts. Model providers are under a Business Associate Agreement and do not use your data to train their models.
It's included on every plan except the standalone Reviews & Messaging tier. If you're already a customer, it's in your dashboard now. There's nothing to install.
Connect your practice management software and the assistant answers from your own data. No report building, no spreadsheet — and any answer downloads as a PDF.
