Question 1

Is the ambient listening feature HIPAA compliant?

Accepted Answer

Yes, the entire infrastructure is HIPAA compliant and HITRUST certified, and ambient listening specifically is built so audio isn't retained permanently unless a patient has consented to that. The voice intelligence pipeline processes the conversation, tokenizes the speech, and extracts the clinical content needed for the note, then the raw audio is handled according to your retention policy rather than sitting in storage indefinitely by default. This matters most for outpatient clinic documentation and telehealth session transcription, where a full day of patient conversations could otherwise represent a significant compliance exposure if audio piled up unmanaged. Every access to stored audio or transcripts is logged, which is part of what the HITRUST certification actually audits.

Question 2

Does the bot work for specialized medical fields like Cardiology or Neurology?

Accepted Answer

Yes, it's trained on specialty-specific medical corpuses, so the terminology and structure it uses for a cardiology encounter looks different from a neurology one, rather than forcing every specialty into the same generic template. This is what the Template Library feature is actually doing under the hood: it adapts documentation structure to your specialty's conventions during the Setup Integration step, so a hospital running rounds across multiple departments doesn't get one-size-fits-all notes. Surgical procedure coding in particular benefits from this since operative notes have very different structural requirements than an outpatient visit note. The specialty templates keep improving as more encounters in that field get reviewed and corrected by physicians.

Question 3

How accurate is the medical coding suggestion?

Accepted Answer

We measure 99.2% accuracy on ICD-10 coding across our client base, and CPT code suggestions follow the same approach: the model reads the full clinical context of the generated note, weighing symptoms discussed in depth more heavily than ones mentioned only in passing, before suggesting a code. That context-driven approach is what keeps accuracy high for surgical procedure coding especially, where the correct CPT code often depends on details buried in the middle of a note rather than the chief complaint at the top. Every suggested code still goes through physician review before submission, since coding accuracy on our end doesn't replace the physician's final sign-off, it just means fewer corrections are needed during that review step.

Question 4

Do patients need to consent to being recorded?

Accepted Answer

Yes, standard practice is to get verbal or written consent before ambient listening starts recording an encounter, and this is a workflow step we build into onboarding rather than something clinics have to bolt on separately. For telehealth session transcription this usually happens at the start of the video call, and for in-person outpatient visits it's typically part of the intake process. We don't record without that consent step being completed first. Clinics running high encounter volume, like hospital rounds, often standardize the consent language once across their EHR intake forms so front-desk and nursing staff aren't manually asking for verbal consent on every single patient throughout the day.

Question 5

How long does it take for the note to be generated after the visit?

Accepted Answer

Notes are typically ready for physician review within 30 to 60 seconds of the encounter ending. That speed is what drives most of the roughly two hours saved per physician per day, since the note is essentially waiting in the queue by the time a physician finishes with the patient and turns to their screen, rather than the physician needing to sit down separately after hours to write it from memory. The Generate Notes step happens automatically right after the recorded encounter closes, using the speech-to-text tokenization and coding extraction that ran in the background during the visit. Physicians still review and edit before anything gets submitted to the EHR; the 30 to 60 second figure is generation time, not total turnaround.

Question 6

Can it filter out casual conversation from the medical note?

Accepted Answer

Yes, the model is trained to separate clinical content that belongs in a SOAP note from rapport-building small talk, like a comment about the weather or asking how someone's kids are doing, so the generated note stays clean and doesn't need physicians to manually strip out irrelevant lines. This distinction happens as part of the same ambient listening pipeline that extracts ICD-10 and CPT coding terms; if a sentence doesn't carry clinical meaning, it doesn't make it into the structured summary. This matters more than it sounds for hospital rounds and outpatient visits where physicians build rapport deliberately, since a note padded with conversational filler would actually take longer to review than one written by hand.

Question 7

Does it integrate directly into Epic or Cerner?

Accepted Answer

Yes, we support SMART on FHIR integrations with both Epic and Cerner, which lets the bot push a completed note directly into the patient's chart rather than leaving it as a separate document someone has to copy over manually. Athenahealth, DrChrono, and Allscripts are supported the same way through HL7 FHIR. This integration is set up once during the Setup Integration step alongside your specialty templates, and after that, the Review and Submit stage becomes a matter of the physician approving the note inside their normal EHR workflow, with no switching between two separate systems. For hospitals running Epic across multiple departments, this is usually the part of onboarding that takes the most coordination with IT.

AI agent case studies

Explore all 11 AI agents

Enterprise Conversational AI: The 2026 Guide

Book a free AI audit

Enterprise Conversational AI: The 2026 Guide

AI Medical Documentation Bot

What is AI Medical Documentation Bot?

AI Decision Intelligence

Everything you need. Nothing you don't.

Ambient Listening

EHR Integration

Medical Coding

Template Library

HIPAA Compliant

Quality Assurance

Without AI

With AI Medical Documentation Bot

Setup Integration

Record Encounters

Generate Notes

Review & Submit

More agents in the suite

AI Clinical Assistant

The services behind this agent

Generative AI Solutions

Artificial Intelligence & Automation

Full-Stack Development

FAQ: AI Medical Documentation Bot

Let's build your AI system