Across nine states, AdventHealth is significantly easing the workload of its care teams by deploying ChatGPT to automate administrative tasks, resulting in an 80 percent reduction in time spent on administrative tasks. The health system, serving millions of patients annually, recognized that increasing administrative complexity was hindering clinicians and staff from focusing on patient care. Previously, physician advisors dedicated approximately 10 minutes per case to utilization management, a process of chart review, detail identification, and rationale drafting, time now partially reclaimed through AI assistance. “We don’t talk about AI as automation; we talk about time back,” says Rob Purinton, Chief AI Officer at AdventHealth, emphasizing the organization’s focus on empowering its workforce rather than simply automating tasks. This shift aims to improve operational efficiency, expand clinical capacity, and enhance the patient experience.
ChatGPT for Healthcare Streamlines Utilization Management
The health system’s approach prioritizes adoption as the primary outcome, rather than focusing solely on technological implementation, a strategy yielding measurable improvements in efficiency and capacity. This isn’t merely about speeding up existing processes; AdventHealth is actively redesigning workflows to leverage the technology’s capabilities. The organization measures impact using data directly from electronic health records, ensuring objective assessment of time savings rather than relying on subjective estimates. “We can see exactly how many minutes have improved and whether that change is statistically significant.” Beyond utilization management, departments like finance and HR are experiencing similar gains, with AI assisting in document creation, information summarization, and policy conversion into usable formats. This training model uses peer groups to share best practices and prompts specific to their functions, fostering organic growth and consistent usage.
AdventHealth deliberately chose OpenAI for its enterprise-grade infrastructure, prioritizing governance controls and reasoning capabilities over simply seeking a demonstration project. “We chose OpenAI because we weren’t looking for a demo; we were looking for enterprise infrastructure,” Purinton explains. The resulting gains are being framed not as cost savings, but as “time back,” capacity reclaimed for clinicians to focus on patient care and staff to pursue higher-value work, with one physician reportedly able to leave documentation tasks behind and spend evenings with family. “He was leaving work at work,” Purinton says. “He could go home and be present with his family.”
AdventHealth Prioritizes AI Adoption as Operational Metric
AdventHealth is moving beyond exploratory artificial intelligence projects, establishing adoption rates as a key performance indicator across its nine-state healthcare system. Rather than solely focusing on technological innovation, the organization is actively measuring how consistently and safely its workforce integrates AI tools into daily routines, a strategy driven by the need to manage increasing demands alongside tight financial margins. This emphasis on practical implementation reflects a broader industry trend of translating AI potential into tangible operational improvements. A significant initial success has been observed in utilization management, where the implementation of ChatGPT for Healthcare is reducing the time spent on portions of the review process. The organization isn’t relying on subjective feedback; instead, it’s tracking measurable data, including timestamps within electronic health records, to confirm improvements. “We prefer measures that are baked right into the process,” Purinton states, ensuring statistically significant results are captured.
This focus on adoption extends beyond clinical applications, with teams in finance, HR, and IT also leveraging AI to streamline documentation and information summarization. The organization measures impact using system-level data, including an 80 percent reduction in time spent on administrative tasks. “We made a decision early on to treat adoption as the product,” Purinton says, highlighting the deliberate shift in focus from simply deploying technology to ensuring its effective utilization. The organization’s choice of OpenAI was also strategic, prioritizing enterprise-grade infrastructure and collaborative innovation over mere demonstration projects, ultimately aiming to return valuable time to clinicians and staff.
The hardest part of AI in healthcare is getting humans to use it safely, consistently, and at scale.
Rob Purinton, Chief AI Officer at AdventHealth
OpenAI Infrastructure Enables Enterprise-Scale Healthcare Deployment
AdventHealth’s ambitious deployment of OpenAI’s ChatGPT isn’t simply a technological upgrade; it represents a fundamental shift in how a large healthcare system approaches artificial intelligence adoption. Rather than isolated pilot programs, the organization, operating across nine states, prioritized enterprise-grade infrastructure to drive consistent and safe AI usage throughout its workforce. This strategic decision stemmed from a recognition that the most significant hurdle wasn’t the technology itself, but rather ensuring widespread, effective implementation. Now, ChatGPT assists with much of this sequence, reducing the time spent on these tasks. Teams are leveraging ChatGPT to generate first drafts of documents, summarize information, and create structured outputs, leading to faster turnaround times and increased capacity. “We don’t talk about AI as automation; we talk about time back,” Purinton emphasizes, highlighting the focus on freeing up staff for higher-value work.
The organization’s approach to measuring adoption, tracking messages per user per business day, treats AI integration as a key performance indicator, monitored and managed with the same rigor as any other operational metric. This commitment to measurable results, coupled with a peer-group training model, is proving crucial to scaling AI across the enterprise, ultimately aiming to return time to clinicians and improve the patient experience.
We had folks who were eager to start, but there were a very large number of people who were on the sidelines.
Rob Purinton, Chief AI Officer at AdventHealth
Reduced Administrative Tasks Return Time to Clinicians
This isn’t simply about streamlining existing processes; the organization is fundamentally redesigning workflows to reclaim valuable time for care teams, a strategy driven by the recognition that operational efficiency directly impacts clinical capacity. This data-driven approach reveals not only time savings but also statistically significant improvements in workflow performance, allowing for targeted reinvestment of capacity. This focus on adoption as a primary outcome shaped the entire rollout. Instead of framing AI as a replacement for staff, AdventHealth positioned it as a tool to alleviate administrative burdens. “We don’t talk about AI as automation; we talk about time back,” Purinton explains, highlighting a deliberate effort to foster buy-in and address concerns about job displacement. The impact is already visible on an individual level; one physician, previously burdened with completing documentation outside of work hours, is now able to finish tasks during the workday and dedicate more time to family. AdventHealth views these initial successes as a foundation for broader applications, including patient access and clinical decision support, all while prioritizing governance, measurement, and trust.
We chose OpenAI because we weren’t looking for a demo. We were looking for enterprise infrastructure.
