Case Study: Humetrix achieves accurate multilingual medical transcription with Speechmatics

Humetrix delivers multilingual medical transcription across 97 countries with Speechmatics

Humetrix, a company that builds multilingual patient intake software for healthcare providers, faced a critical challenge with transcription accuracy. Their original engine, OpenAI's Whisper, produced unacceptably high error rates on complex medical terminology, which posed a direct risk to patient safety. To deliver reliable clinical-grade intake across 27 languages, they needed a more accurate solution from a vendor like Speechmatics.

Speechmatics provided its speech recognition technology, which was integrated into Humetrix's on-premises AWS workflow to transcribe patient interviews in real time. This solution enabled highly accurate multilingual medical transcription, which was successfully proven at the Paris 2024 Olympics, supporting patients from 97 countries. The partnership allowed Humetrix to replace Whisper with Speechmatics, achieving the precision required for safe clinical deployment.

Humetrix