Introduction
DeepScribe is a San Francisco company founded in 2017 that builds an ambient artificial intelligence platform for clinical documentation — capturing patient encounters and generating structured notes clinicians review before sign-off. Matthew Ko serves as chief executive officer and co-founder per DeepScribe’s about page, which states the Ambient Operating System is deployed in more than 1,500 healthcare organizations with over 8 million visits captured and a 99.92% note satisfaction rating cited by the company.
DeepScribe distinguishes itself from general-purpose scribes by emphasizing specialty workflows — particularly community oncology — where care complexity, team coordination, and longitudinal context make generic transcription inadequate. An April 2026 urology webinar recap on DeepScribe’s site documents how the platform handles PSA trajectories, IPSS scoring, cystoscopy documentation, and automated E/M coding in urology-specific language.
What the company does
DeepScribe’s Ambient Operating System listens to clinical conversations with appropriate consent and security controls, then drafts notes mapped to EHR fields using models fine-tuned for specialty vocabulary and provider preferences. SmartPrep pulls prior notes, labs, imaging, and pathology from integrated EHRs into structured pre-charting summaries days before visits — reducing the information-gathering burden on medical assistants and nurses.
Beyond documentation, DeepScribe automates coding workflows: established diagnoses carry forward, new problems discussed during encounters are added, and E/M levels calculate from medical decision-making complexity. When EHR-integrated, HCC coding assistance surfaces eligible codes as conversations address relevant clinical criteria — connecting ambient AI to risk-adjustment revenue integrity.
DeepScribe describes its platform as an ambient operating system extending beyond scribing into revenue cycle and workflow intelligence — a framing that distinguishes specialty ambient vendors from transcript utilities that stop at note text without coding or pre-visit context layers.
Who it serves
DeepScribe serves specialty practices and health systems where documentation quality and coding accuracy directly affect margins — community oncology networks, urology groups, and multi-specialty ambulatory organizations cited on its about page. The company reports serving 2,800+ clinics and health systems with onsite implementation teams for complex care environments.
Buyers typically evaluate DeepScribe on adoption rates — the April 2026 urology webinar states organizations below 80% utilization rarely see meaningful returns, while DeepScribe reports overall adoption above 85% after onboarding that profiles each provider’s documentation style before go-live.
Company background
DeepScribe was founded in 2017 in San Francisco by Matthew Ko and co-founders including Akilesh Bapu, emerging from natural language processing research applied to medical summarization. Ko’s personal motivation — coordinating care during his mother’s breast cancer treatment — appears on the about page as the origin story for building ambient AI that preserves clinician presence. The company has raised venture funding from investors listed on its site and expanded from documentation into what it describes as an ambient operating system spanning pre-charting, coding, and workflow intelligence.
Matthew Ko remains the public CEO voice in DeepScribe leadership content and specialty webinar materials reviewed July 10, 2026.
Product and AI capabilities
DeepScribe combines speech recognition, medical natural language understanding, and generative models tuned for specialty note formats. The April 2026 urology webinar demonstrates context-aware documentation: a mock prostate cancer consult generates SOAP notes with PI-RADS scoring and fusion biopsy planning without clinicians verbalizing every lab value because SmartPrep already incorporated EHR data.
DeepScribe’s adoption methodology — reviewing prior notes to match formatting, terminology, and structure before go-live — addresses a common ambient AI failure mode where first-draft notes feel foreign and clinicians revert to manual entry. The platform learns from subsequent edits to apply preferences automatically.
Key developments
2017: DeepScribe founded in San Francisco. 2020s: Platform scales to 1,500+ organizations per about page metrics. April 2026: Urology ambient AI webinar published documenting SmartPrep, specialty coding, and adoption thresholds. DeepScribe reports 8 million+ visits captured and 99.92% note satisfaction rating on about page.
The April 2026 webinar quotes DeepScribe directors Blake Swinehart and Ravneet Riar on urology-specific language — PSA trends, LUTS scoring, urodynamics, and cystoscopy documentation — that general ambient models mishandle. Mock encounter content shows E/M coding generated without clinicians spelling out every imaging finding because SmartPrep pre-incorporated EHR data.
Why it matters
DeepScribe matters because ambient AI vendors often win primary-care pilots but struggle in subspecialty workflows where vocabulary and longitudinal context differ materially visit to visit. DeepScribe’s urology and oncology focus tests whether specialty-tuned models improve coding capture and clinician retention in high-complexity ambulatory care. Compare with enterprise ambient vendors Abridge and Nabla scaling across health-system service lines with different EHR integration priorities.
DeepScribe’s stated 85%+ adoption threshold and note-style matching before go-live address a recurring enterprise objection: ambient AI that produces alien note formats dies in pilot phase regardless of transcription accuracy. Community oncology networks named on the about page represent a buyer segment where documentation burden directly affects physician supply in underserved markets.
Sector context
DeepScribe belongs in Brel’s AI in healthcare coverage under ambient clinical documentation and specialty workflow AI, alongside Abridge, Ambience Healthcare, and Nabla. Oncology-focused precision medicine platforms such as Tempus AI operate downstream of documentation in molecular data workflows — adjacent but distinct buyer committees.
Sources and references
DeepScribe urology webinar recap (April 2026) and about page with deployment metrics.
- DeepScribe — Urology ambient AI webinar recap (deepscribe.ai, April 2026)
- DeepScribe — About (deepscribe.ai)
Reviewed July 10, 2026. Matthew Ko is CEO and co-founder per about page. Founded 2017 in San Francisco.