Ambient AI Medical Scribe

Dictum listens to patient encounters in real time and produces structured clinical notes for clinician review. Instead of typing during the visit or reconstructing notes afterward, you get a draft that captures the conversation's clinical content — ready for your edits before it enters the EHR.

Start Free with Dictum

Clinicians should review AI-generated documentation before adding it to the medical record and should use Dictum in accordance with their organization's policies and applicable laws.

What is an ambient AI medical scribe?

An ambient AI medical scribe records the natural conversation between a clinician and a patient during an encounter, then converts that audio into a structured clinical note. The word “ambient” means the tool works in the background — there is no need to dictate into a microphone using specific commands or navigate a template while the patient is in the room.

The goal is straightforward: reduce the time clinicians spend on documentation without losing the clinical detail that good notes require. Rather than replacing the clinician’s judgment, an ambient scribe produces a starting draft. The clinician reviews that draft, corrects anything inaccurate, adds context the AI could not infer, and approves the final version. This is fundamentally different from a human scribe who edits in real time — the AI version gives you a complete draft after the encounter ends, and you decide what stays.

How Dictum captures patient encounters

When you start a session in Dictum, the app begins recording the encounter audio through your device’s microphone. You can use a phone, tablet, or laptop — whatever you normally have in the exam room. The audio is streamed and processed in real time, so there is no large file sitting on a server when the visit ends.

During the conversation, Dictum distinguishes between speakers and tracks the flow of the clinical discussion. It identifies segments that correspond to history-taking, review of systems, physical exam findings discussed aloud, assessment reasoning, and plan items. This isn’t keyword matching — the system uses medical language models trained on clinical conversation patterns.

Once you end the recording, Dictum generates a note draft. The turnaround is typically within one to two minutes, depending on the length of the encounter. You then review the note, make edits, and export or copy the final version to your EHR.

Ambient capture workflow

  1. Patient encounter
  2. Ambient capture
  3. AI-generated draft
  4. Clinician review
  5. EHR-ready note

From conversation to SOAP note

The most common output format is a SOAP note — Subjective, Objective, Assessment, and Plan. Dictum maps conversation content to each section based on clinical context rather than rigid keyword rules.

Patient-reported symptoms, medication history, and concerns go into the Subjective section. When you discuss exam findings, vitals, or lab results aloud, Dictum places those in Objective. Your clinical reasoning — differential diagnoses, impressions — populates Assessment. Treatment decisions, referrals, prescriptions, and follow-up instructions fill the Plan.

This mapping is not perfect every time. Complex encounters with multiple problems, interruptions, or non-linear conversation flow may produce drafts that need reorganization. That is expected, and it is why clinician review is a required step, not an optional one. You can learn more about how the note structure works on our AI SOAP note generator page.

When ambient capture is useful

Ambient scribing works best when the clinician is having a spoken conversation with the patient and wants to remain present during the visit rather than documenting simultaneously. Common scenarios include:

  • Primary care and internal medicine visits where history-taking and counseling make up most of the encounter.
  • Behavioral health and psychiatry sessions where typing during the appointment can feel intrusive to the patient and disrupt rapport.
  • Specialty consultations that involve extended discussions of imaging, pathology, or treatment options.
  • High-volume clinic days when documentation backlog can push charting into evening hours.
  • New patient encounters that involve detailed histories and multi-system review.

In each case, the common thread is that the clinician would otherwise need to reconstruct the visit from memory or brief shorthand notes after the patient leaves. Ambient capture reduces that reconstruction effort by giving you a draft that already reflects the conversation.

When post-visit dictation may be better

Ambient capture is not the right tool for every situation. If the encounter involves sensitive discussions that the patient prefers not to have recorded, dictating your note after the visit gives you full control over what goes into the documentation. Some clinicians also prefer dictation when the encounter is primarily procedural with little conversation — for example, a quick wound check or suture removal.

Dictum supports both approaches. You can use the post-visit dictation workflow to record a summary in your own words after the patient leaves. Many clinicians use ambient capture for most visits and switch to dictation for shorter or more sensitive encounters. The two workflows produce the same structured output — the difference is what audio goes in.

Security and clinician review

Every note Dictum generates is a draft. It is not automatically added to any medical record, and it does not leave Dictum unless you explicitly export or copy it. Clinicians should review AI-generated documentation before adding it to the medical record and should use Dictum in accordance with their organization’s policies and applicable laws.

On the privacy side, Dictum processes audio in real time and does not retain patient audio recordings on our servers after transcription. Data is encrypted in transit using TLS 1.3 and at rest using AES-256. We offer Business Associate Agreements for organizations that require them. For full details, see our HIPAA-focused documentation workflows page.

Supported clinical workflows

Dictum produces review-ready documentation in multiple formats. SOAP notes are the default, but the system supports other structures depending on specialty and preference:

  • SOAP notes — standard four-section format used across most outpatient settings.
  • H&P notes — history and physical format common in initial consultations and hospital admissions.
  • Progress notes — focused follow-up documentation for ongoing care.
  • Procedure notes — structured documentation for minor in-office procedures.
  • Patient instructions — plain-language summaries of the plan that can be shared with the patient.

Each output type can be adapted to your specialty. Dictum includes templates for over a dozen specialties, from family medicine to orthopedics. Check Dictum pricing plans to see which note types are available on each tier.

Frequently asked questions

Dictum records the audio of the clinical conversation between the clinician and the patient. The audio is processed in real time and is not stored on Dictum servers after transcription. The resulting transcript is used to generate a structured clinical note draft for the clinician to review.

Consent requirements vary by state, organization, and practice setting. Many jurisdictions require that patients be informed when a conversation is being recorded. Clinicians should follow their organization's policies for patient notification and consent before using any ambient capture tool.

Yes. Every note Dictum produces is a draft meant for clinician review and editing. You can modify any section, add clinical detail, remove inaccurate content, or restructure the note before adding it to the patient's medical record.

Accuracy depends on factors like microphone quality, background noise, speaker clarity, and the use of specialized medical terminology. Dictum uses medical-vocabulary-aware speech recognition, but no transcription system is perfectly accurate. Clinician review of every generated note is required before EHR entry.

Dictum can capture conversations in typical outpatient exam rooms and office settings. Very noisy environments, such as busy emergency departments or shared spaces without barriers, may reduce transcription quality. Using a directional or lapel microphone can help in louder settings.

Dictum is built with HIPAA-focused workflows. Audio is processed in real time and not retained on our servers after transcription. Data is encrypted in transit and at rest, and Dictum offers Business Associate Agreements for covered entities. See our HIPAA compliance page for full details.

Try ambient AI scribing with Dictum

Record your next patient encounter and get a structured clinical note draft in minutes. Free to start.

Try Dictum Free