A global healthcare company sought to improve speech detection algorithms used by health care professionals (HCPs) to access patient information.
A global healthcare company needed to increase efficiency and accuracy of accessing patient information via both voice and text input.
We were asked to explore how health care professionals would naturally interact with a text and voice-controlled system in order to access information about their patients.
We interviewed more than 500 health care professionals from the US, UK, and Ireland, ensuring variety in our sample collection.
To learn how HCPs might interact with the system, we asked them to naturally request patient data in their own words, collecting both text and voice samples of various scenarios for each participant.
The diversity of our participants allowed us to capture a wide range of text and voice samples. We learned how HCPs would ask a machine for information in both text and voice formats and collected a large library of regional accents to use in training language algorithms to increase efficiency in these interactions.