Dive Brief:
- Microsoft and Mercy entered a long-term partnership to deploy generative AI-based solutions across the health system, the two companies said Wednesday. The healthcare organization will use Azure OpenAI Services tools to assist patients, improve scheduling procedures and power an administrative chatbot.
- “With Microsoft, we are exploring more than four dozen uses of AI and will launch multiple new AI use cases by the middle of next year to transform care and experiences for patients and coworkers,” Joe Kelly, Mercy's EVP of transformation and business development officer, said in the announcement.
- Microsoft is leveraging its partnership with OpenAI to grow its healthcare sector cloud footprint. Last month, the tech giant stood up an electronic health records cloud database solution with Epic, and the two companies are working to infuse Epic’s EHR systems with generative AI.
Dive Insight:
Chatbots, administrative assistants and customer-facing automation are three of generative AI’s more immediate applications. Driving cloud business expansion is another clear consequence of the technology and its enterprise adoption.
All of the major cloud providers, including Microsoft, AWS and Google Cloud, have raced to rapidly deploy off-the-shelf models on their platforms, anticipating an AI boom to accelerate revenue growth.
The heavily regulated, risk-averse healthcare sector represents a massive yet challenging opportunity for cloud expansion, as well as for generative AI tools.
Mercy’s initial plans for the technology are in line with use cases piloted in other sectors — ingesting and summarizing large documents, improving call center operations and enhancing chatbot capabilities.
Additional use cases are in development, according to Wednesday’s announcement, following a hackathon that brought together engineering teams from Mercy and Microsoft.
Protective guardrails will need to be in place before widespread deployments in the healthcare sector proceed. Introducing untested technologies in areas related to patient safety and privacy requires a reliability and accuracy that generative AI tools have yet to demonstrate consistently.
“There's a lot of work to do,” Peter Fleischut, group SVP, chief information and transformation officer at NewYork-Presbyterian, said during a CIO Dive Live event Sept. 13.
“It’s critical that we go through and have the right processes with the right architecture to be able to do this, to make sure that the patient and the patient’s data is safe,” said Fleischut.
