BrazilHealthcare and life sciences2026 – ongoingHapvida
In-house voice AI for consultation transcription, anamnesis and ICD code suggestion
Hapvida's in-house voice AI transcribed 604,383 consultations in five months, with physicians agreeing with the suggested code 94% of the time
Problem
Physicians spent consultation time writing records. Hapvida wanted to return that time to patient care.
Approach
Hapvida built a voice transcription system that listens to consultations in real time, drafts the anamnesis and suggests a CID (ICD) code. Physicians review and validate everything before the record is final. It ran in telemedicine nationwide and in all in-person emergency units in Ceará, with expansion to other states planned by September 2026.
Results
Takeaway
All figures come from Hapvida. The physician stays the final reviewer of every record.
Categories
- Size
- Large enterprise
- Industry
- Healthcare and life sciences
- Business function
- Production and qualityInternal work and knowledge management
- Innovation type
- Process innovation
- AI technology
- Generative AI
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
The article reports on Hapvida's in-house voice AI. It gives the number of consultations transcribed from January to May 2026, physician agreement with the suggested code, audit approval of anamneses and the rise in recorded information.
Full details are in the original sources.
