Sunoh.ai

The note is the model's output. Ambient speech recognition and summarization turn the conversation into a transcript, a dialogue flow, a draft note sorted into Progress Note sections and draft orders, and every output comes from those models. Sunoh is sold as its own product at its own per user price rather than folded into a record system contract; what it takes from eClinicalWorks is distribution.

Everything Sunoh produces stays a draft until the clinician acts on it. The provider reviews the summarized content, edits it and imports it into the Progress Note, merging preset defaults with one click, and through EHR Sync reviews, edits and approves notes before they reach another record system.

The draft goes past the note. Sunoh captures labs, imaging, procedures, medication orders and follow up details from the conversation and pre fills them as orders, so a proposed clinical action is placed in front of the clinician already formed. Nothing published describes what the clinician confirms for an order specifically, how an uncertain capture is flagged, or how accurate order capture is.

The engine is described as ambient listening, voice recognition, natural language processing and machine learning, and Sunoh says the documentation it produces is highly accurate.

The model, its provider and any accuracy figure are not published.

Microsoft Azure is the one party named. Sunoh's pricing page says it runs on Azure, and eClinicalWorks states that its ancillary products run on the eClinicalWorks cloud in Azure data centers across several US regions. Above hosting nothing is named: no speech recognition engine, language model or model provider, and no subprocessor list. Nothing says whether a model service inside or outside Azure generates the note.

Ownership shapes where data can go. Sunoh is built inside the eClinicalWorks group, and healow's terms allow client data to be used in de identified form for data benchmarking, sharing, warehousing and similar analysis services.

Evidence is customer testimony. Sunoh.ai publishes more than 30 customer stories from practices and health centers such as Family Health Centers, Hendry Regional Medical Center, Indiana University Health Center and MedFlorida Medical Centers. Their quotes report up to two hours saved a day, notes closed the same day rather than within three days, and one practice cutting documentation time by 50 percent or more.

No peer reviewed study, controlled comparison, independent evaluation or accuracy benchmark is published, and the figures carry no stated method.

Sunoh states that it uses industry standard encryption and security protocols and includes a business associate agreement with every plan, and healow's privacy policy says protected health information it processes for providers is governed by the customer agreements. healow's terms allow client data to be used in de identified form for benchmarking, sharing, warehousing and similar analysis.

How long encounter audio and transcripts are kept, and whether they train models, is not published.

A business associate agreement is listed as included in the plan on Sunoh's pricing page, so it comes with the standard purchase rather than on request. healow's privacy policy has a section on healow acting as a business associate, under which protected health information is processed according to the service agreements and business associate agreements providers sign, and healow's terms say those customer documents prevail where they conflict with healow's standard terms. The parent company reports a completed third party HIPAA compliance audit.

The agreement itself is not published, so its terms, including how it treats audio and transcripts, cannot be read before sign up.

The certifications sit with the parent company. Sunoh is a healow product, and healow is an eClinicalWorks company. eClinicalWorks states that its ancillary products run on the eClinicalWorks cloud in Microsoft Azure data centers across several US regions, and lists HITRUST, ISO 27001, PCI DSS, CSA STAR Level 1 and Level 2, TX-RAMP and a completed third party HIPAA compliance audit.

Whether those certifications cover Sunoh's ambient capture and note generation, as distinct from the record system, is not stated. Azure's own attestations cover the hosting layer rather than the vendor's controls.

Sunoh.ai drafts notes and orders for a clinician to edit and import, and healow's terms exclude healthcare services from what it provides; the product holds no FDA clearance.

Two features go past narrative documentation. Sunoh pre fills orders for labs, imaging, procedures and medications from the conversation, which puts a proposed clinical action in front of the clinician already formed. It also documents encounters held in Spanish, Portuguese, Cantonese, Mandarin, Vietnamese and other languages, a translation step that a clinician who does not speak the patient's language cannot check. Neither feature carries a published statement of intended use.

Multilingual documentation is one of Sunoh's main selling points. It markets encounters documented in English, several dialects of Spanish, Portuguese, Cantonese, Mandarin, Vietnamese and other languages, and healow has spoken publicly about multilingual scribes letting patients speak in their preferred language. Errors in speech recognition and translation can compound, and the patients most exposed are the ones the feature is marketed to serve.

Accuracy by language, accent or dialect is not published.

Review before import is the mechanism for catching an error: the clinician edits the draft and approves what reaches the record, in eClinicalWorks or through EHR Sync. healow's terms, updated 1 June 2026, carry a disclaimer and a limitation of liability, send disputes to arbitration with a class action waiver, and give way to the customer's own service agreement and business associate agreement where those conflict. The contract and support path often runs through eClinicalWorks, since many practices reach Sunoh inside that record system.

No accuracy figure, error rate or warranty accompanies the output.

Sunoh is native to eClinicalWorks: it runs inside the desktop record, eClinicalTouch on iPad and eClinicalMobile, imports structured content straight into the Progress Note and pre fills orders. Beyond its home system, Sunoh.ai names integrations with athenahealth, Cerner, Epic, ModMed, Practice Fusion and Tebra, without published detail on what each reads or writes.

Other browser based record systems are reached through EHR Sync, a free Chrome extension configured once per EHR, which reads patient data, drafts notes and sends them into the record after the clinician reviews and approves them. How the extension authenticates, what it can access and how its writes appear in the host system's audit log are not described.

Hosting is named; the region is not. Sunoh's pricing page says it runs on Microsoft Azure, and eClinicalWorks states that its cloud uses Azure data centers in several US regions, with each client database replicated to a failover site in another region. That describes the parent's cloud rather than a residency commitment for Sunoh, and nothing says where inference runs.

Clinicians use Sunoh on iPhone, iPad, Android and the web, natively inside eClinicalWorks, or through the EHR Sync extension in other browser based records, so what a deployment looks like depends on which record system a practice runs.

The price is on the site. Sunoh.ai lists $149 per user per month, marked as a limited time rate, with a sign up that carries no contract commitment and with implementation services, the voice recognition system and a business associate agreement included. EHR Sync carries no extra charge, so the cost can be sized and the product started without a sales conversation, though the page does not say what the rate becomes when the offer ends.

Specialty reach is listed page by page. Sunoh.ai publishes pages for family and internal medicine, pediatrics, cardiology, dermatology, orthopedics, urgent care, ambulatory surgery centers, behavioral health with DAP notes, vision clinics, and dental practices, where it produces periodontal charting and treatment plans. Dental and optometry documentation follow different structures from physician notes, which widens the fit for a multispecialty or dental group.

Encounters can be documented in English, several dialects of Spanish, Portuguese, Cantonese, Mandarin, Vietnamese and other languages, and pediatric visits with several speakers, including a parent or an interpreter, are captured. Performance by language is not published.