Pharmacy-tuned AI Scribe — listens, drafts, codes.
Tap record. Have the conversation. The encounter is captured, transcribed, summarized into a SOAP note, and pre-coded with the right CPT/HCPCS — pharmacist-reviewed in under 90 seconds. Built for clinical pharmacy, not adapted from a primary-care tool.
Three taps. Done before the patient is back in their car.
The pharmacist stays in the conversation. The Scribe captures everything else.
Tap record
Scribe runs on the in-store iPad or the pharmacist's iPhone. With patient consent, it captures audio for the duration of the encounter — no cloud streaming, no third-party microphone access.
Have the conversation
The pharmacist counsels the way they always have. The Scribe transcribes in real time, identifies speakers, and tags clinical entities — meds, doses, complaints, vitals, lab values.
Review and sign
A complete SOAP note appears in the chart with suggested CPT/HCPCS code and time documentation. Pharmacist reviews, edits if needed, signs. Ready to bill.
Patient reports improved BP self-monitoring; concerned about persistent dry cough. Denies chest pain or SOB. Adherent to lisinopril; missed metformin doses 2× last week per own report.
BP today 138/86. Home avg (30d) 142/88. Last A1C 7.4%. eGFR 52. PDC metformin 71%, lisinopril 92%.
DM2 — A1C above goal. Likely lisinopril-induced cough; uncontrolled HTN. Adherence concern with metformin.
Recommend MD-collab note re: ARB switch. Reinforce metformin adherence; pillbox + alarm. F/U 2 weeks.
Trained on pharmacy encounters, not primary-care visits.
Most AI scribes were built for 30-minute MD office visits. They mis-handle the dense, medication-centric vocabulary of pharmacy work — drug names, doses, brands vs. generics, adherence concepts, the difference between PDC and MPR. We retrained from the ground up on de-identified pharmacy encounter data.
- Drug name disambiguation across 4,800+ NDC entities (Jardiance vs. Januvia vs. Janumet)
- Dose, frequency, and route capture — including patient-described regimens that don't match the script
- Adherence flags (missed doses, hoarding, sharing, side-effect avoidance)
- Vitals capture (BP, HR, weight, blood glucose, peak flow) including patient-reported home readings
- Distinct templates per service: CCM, MTM CMR/TMR, immunization, transitions, PrEP, smoking, contraception
A scribe you can sign your name on.
The pharmacist always reviews before signing. We measure accuracy on a validated test set the same way we'd measure a clinical decision support tool.
What 90 seconds saved per encounter looks like across a week.
For a pharmacist running 18 clinical encounters a day, 90 seconds saved per encounter is 27 minutes back. Across a week, that's 2 hours of patient-facing time recovered.
A pharmacist's documentation day
Same pharmacist, same week
Aggregate from 168 pharmacists, 90 days post-rollout, vs. their own pre-rollout baseline.
The questions every pharmacy operator asks before turning Scribe on.
Is this HIPAA-compliant? Where does the audio go?
Yes. Audio is encrypted in transit and at rest, processed inside our HIPAA-aligned AWS environment, and deleted within 24 hours of note completion (configurable per pharmacy). A BAA is in place before any PHI moves through the system. Read the security overview and HIPAA documentation.
Do you use my pharmacy's data to train shared AI models?
No. Per-tenant data isolation is hard-enforced. We never use one customer's PHI to train models that serve another customer. We do train tenant-specific personalization models (vocabulary, common drugs, common visit types) and you can opt out at any time.
What happens when the AI gets something wrong?
The pharmacist always reviews the draft before signing. Common error modes — drug-name misrecognition, dose misinterpretation, ambiguous patient statements — are surfaced as flagged fields in the review screen, not silently committed. Each correction the pharmacist makes feeds the per-tenant fine-tuning loop, so accuracy improves on your specific patient population over time.
What about patient consent?
The Scribe is opt-in by encounter. The first time a patient is recorded, MedMe captures a consent attestation (verbal or written) that is logged in the chart. Patients can decline, and the pharmacist can chart the encounter the traditional way without leaving the workflow.
Does Scribe handle Spanish encounters?
Yes — Scribe is bilingual English/Spanish out of the box, with code-switching support (the patient speaks Spanish, the pharmacist responds in English, both are captured cleanly). The note is always rendered in English unless your pharmacy has elected Spanish charting.
What if my store has poor cellular reception?
Scribe runs on Wi-Fi or LTE, falls back to local-buffered mode with no internet, and syncs when connectivity returns. The encounter never disappears — it just gets transcribed when the device is back online.
How does this interact with the billing module?
The suggested CPT/HCPCS code, time elapsed, and modifiers flow directly into the billing console. The pharmacist's signature on the note triggers the eligibility check, code validation, and clearinghouse submission — no second screen, no copy-paste.
End pajama-time charting. Sign every note before the patient leaves.
15-minute live demo on a sample CCM encounter. We'll record, draft, code, and sign — start to finish.