Summary
Sonata Software transformed a US healthcare service platform's manual documentation process by implementing a Gen AI-powered intelligent transcription solution. Using Azure OpenAI technology, the solution automatically transcribes physician-patient conversations in real-time and structures them into standardized electronic health record (EHR) sections with appropriate medical coding. This automation now manages 20,000+ patient encounters monthly while delivering $350,000 in annual savings.
Client Overview
A dermatology physician services organization focused on forming partnerships by acquiring growth‑focused dermatology groups. It provides a single, integrated platform that lightens administrative load and reduces costs across every business function of a busy medical group.
Pressure Points
Heavy reliance on 50+ manual scribes with rising market rates
Each scribe handling 25+ patient records daily at approximately $25/hour
Extended 6-8 week hiring and training cycles for new scribes
High administrative burden on operational resources
Scalability challenges in meeting growing documentation demands
Solutions
Sonata leveraged its gen AI expertise to develop an intelligent automation solution that was implemented to address the client's operational challenges. The solution features a gen AI-driven transcription plugin that captures physician-patient conversations and, using Azure OpenAI, processes these interactions into structured EHR sections such as impressions, morphology and plans. Additionally, it automatically assigns relevant medical codes like international classification of diseases and current procedural terminology (CPT), ensuring alignment with healthcare compliance standards and reducing the administrative load on physicians.
iPad Power App enables start/pause/stop recording and displays live transcripts with keyword prompts; access requires a corporate UDP email ID
Transcript view can be disabled once stable; all practitioner edits are logged for traceability
Solution performs text matches on the transcript. If the practitioner conversation aligns to keywords in electronic medical assistant (EMA), it will result in better quality of match
Save transcript, keywords to blob storage for audit purposes
Results that Speak Volumes
20,000+ patient encounters managed monthly through Gen AI
~$350,000 annual savings in operational costs
Decrease in hiring/ training overheads for manual scribes
