## TL;DR
AI is fixing the EHR problem -- ambient scribes that listen to doctor-patient conversations and write clinical notes automatically, saving physicians 2+ hours/day on documentation and reducing burnout by 40%. From Nuance DAX to Epic, AI transforms EHR from a burden to an assistive partner.

## Core Explanation
EHR AI: (1) Ambient scribe -- microphone records patient encounter; ASR transcribes; LLM extracts medical facts and generates structured SOAP note. Integration into EHR (Epic, Cerner); (2) Clinical NLP -- extract diagnoses (ICD-10), medications, allergies, lab results from free-text clinical notes. NER specialized for medical entities; (3) Automated coding -- AI assigns billing codes from documentation. NLP + rule-based + ML; (4) In-basket management -- GPT-generated draft responses to patient messages, reviewed by physician.

## Detailed Analysis
Nuance DAX Copilot (2023, Microsoft acquisition $19.7B): ambient AI for clinical documentation. Multi-speaker diarization separates doctor/patient speech. SOAP note generation with structured sections. Kaiser Permanente study (2024): 50-70% reduction in documentation time, 79% of physicians report improved work-life balance. Abridge (2018-2025): academic medical center focus. Real-time patient-friendly summaries alongside clinical notes. Epic (2024): integrated ambient AI + GPT-4 in-basket message drafting. Deployed at 500+ hospitals within months. Amazon Comprehend Medical: pre-trained NER for protected health information (PHI), medication, and medical condition extraction. Key impact: physician burnout -- 50%+ of US physicians report burnout symptoms. A major driver: "pajama time" -- 1-2 hours of evening EHR documentation. AI ambient scribes directly address this.