The Beat How Mary Varghese Presti and Microsoft Are Using AI Agents to Give Nurses Their Time Back
About Mary Varghese Presti:
Mary Varghese Presti is a transformational healthcare leader with over two decades of experience spanning clinical care, federal reform, biopharma, and health technology. As Corporate Vice President of Microsoft Health & Life Sciences, she drives growth in complex environments by creating clear strategy, aligning organizations, and operationalizing execution with discipline. Her prior roles include leading Nuance’s Dragon Medical business, overseeing IBM Watson Health’s Life Sciences portfolio, incubating new ventures at athenahealth, and driving digital-health transformation at Pfizer. She began her career as a pediatric nurse at Johns Hopkins and later helped shape national health IT and payment reforms at Booz Allen. Known for navigating complexity with optimism and rigor, she consistently turns ambiguity into strategy and strategy into measurable results.
Things You’ll Learn:
- AI in healthcare is evolving from simple assistants to agentic services that can independently execute predictable workflows, allowing clinicians to regain time and focus. This shift enables a hybrid workforce where human and digital colleagues work side by side.
- Dragon Copilot for nurses was designed specifically to support the way nurses document care, capturing structured inputs such as vitals, intake/output, and observations through natural speech. By reducing EHR time and ambiently recording bedside interactions, it helps turn “caring out loud” into complete documentation.
- Nurses spend more than a quarter of their 12-hour shifts documenting in the EHR, often feeling emotionally torn between screens and patients. AI that listens in the background can significantly reduce this burden while allowing for more presence at the bedside.
- New tools are starting to expose the “invisible work” nurses perform, from constant micro-assessments to coordination with ancillary departments. Making this work visible is a critical step toward properly valuing nursing labor and improving workforce planning.
- Real-world use cases, such as AI agents assembling data for tumor boards at academic centers, show that agentic workflows can compress decision timelines from weeks to days. These same principles can be extended to many clinical and non-clinical tasks, accelerating care while preserving clinician judgment.
Resources: