
Seventy minutes is enough time to complete a colonoscopy or a hernia repair; it’s also the average time saved per patient visit when healthcare providers use fully automated administrative workflows that not just collect, but also act on data.
Clinicians are drowning in alerts, forms, and follow-ups, while $90 billion per year is spent on routine administrative tasks. Agentic AI brings a breakthrough, taking these informative alerts, and making decisions based on their context.
In healthcare marketing, we are already seeing agentic AI independently plan campaigns, test variations, adjust budgets, and launch A/B tests without human intervention. Within two months, a mid-sized healthcare IT company tripled its lead conversion rate, and the cost per lead dropped by 40%.
These tools will be especially helpful for clinicians overwhelmed by notifications by handling tasks like automating follow-ups, matching patients with the right doctor, and checking details such as insurance coverage. The result is shorter task lists and more streamlined administrative processes that enable medical practitioners to give more hands-on time to their patients.
Broken administrative workflows costs
Healthcare organizations, in particular, suffered from the COVID-19 pandemic, and these institutions continue to face rising costs to attract and retain essential staff. Over the next decade, about 194,500 openings for registered nurses are projected each year. However, faculty shortages also mean that not enough nurses can be trained in time.
A patient’s interaction with a healthcare provider often begins long before they see a doctor and continues well after. From booking appointments to settling bills, the administrative process is a crucial part of care, but it also consumes a substantial amount of medical teams’ time.

Source: The 2024 CAQH Index
The 2024 CAQH Index investigated multiple providers and health plans to understand whether administrative tasks were fully electronic, partially electronic, or manual, the transaction volume, and the cost and time required to complete them. Key findings include:
- $90 billion: Annual cost of routine administrative tasks like insurance checks.
- $20 billion: Potential savings from shifting to electronic workflows.
- 70 minutes: Average time saved per patient visit with fully automated administrative processes.
Minimizing manual tasks not only cuts costs and saves time, but also has the potential to enhance staff satisfaction and improve overall patient care.
Where agentic AI can help
Over a third of healthcare workers continue to struggle with being overworked, with burnout levels still higher than pre-pandemic. Agentic AI is here to assist medical practitioners with shrinking their task lists and giving them a better work-life balance.
For example, agentic AI can be safely used to make process-based and rule-bound decisions, such as scheduling follow-ups or general appointments. Since many follow-up or appointment decisions are based on established clinical guidelines — such as touching base with diabetic patients every 3 months or scheduling consults within 2 weeks if LDL cholesterol is above a certain threshold — agentic AI can be programmed with these rules and timelines. They can make decisions that align with clinical standards, without improvising or overriding medical judgment.
They can also help with filling out certain forms based on structured data. For a referral form, the AI can automatically extract, match, and insert the patient’s latest diagnosis, primary physician information, and reason for referral based on notes or codes already entered in the EHR.
We are beginning to look at what clinical workflow orchestration AI can assist with as well, where it can act within defined boundaries, such as triggering care pathways after certain events. So, if a patient is discharged from the hospital after treatment for heart failure, as soon as the discharge note is entered into the EHR, the AI detects this event. It can then schedule a follow-up cardiology visit within 7 days, and notify the patient with a personalized SMS:
“Hi Alex, don’t forget your cardiology appointment on Wednesday at 8:00 AM. Let us know if you need help getting there.”
The patient’s response can trigger the AI to book transportation for the patient if needed.
There are other scenarios where AI agents can assist or act, but they only submit the action once verified by an expert. Some use cases here include preparing treatment instructions or drafting non-standardized follow-up messages to patients.
What healthcare providers can do to prepare
Although agentic AI’s advantage is its ability to act, the integrity of the healthcare industry depends on its clinicians. They must oversee these tools to reap the benefits.
Healthcare providers who empower their teams to lead AI-driven care can safely leverage these medical assistants. That means involving doctors, nurses, and frontline staff in the development process of AI agents. Medical teams are far more likely to trust and adopt AI practices properly if they helped build them or choose what they agree would be helpful use cases.
The healthcare industry is at an inflection point as understaffed and exhausted teams struggle to keep up with rising demands. Agentic AI presents a solution, and those who get involved early will have a better learning experience. Some best practices for healthcare providers when choosing solutions include:
- Investing in AI tools that can integrate with existing administration systems.
- Validating HIPAA compliance and implementing internal governance policies.
- Training teams on how to navigate AI medical assistants and best leverage these tools.
- Ensuring medical practitioners remain in the driver’s seat at critical touchpoints.
- Empowering medical professionals to lead AI-driven care confidently.
Medical professionals and their health tech partners must serve as AI’s ethical compass. By guiding strategy, refining communication, and upholding governance standards, healthcare teams can protect the integrity of the field, while also delivering relief for our most essential staff.
About Nate MacLeitch
Nate MacLeitch is a highly experienced business professional with a diverse background in industries such as telecom, media, software, and technology. He began his career as a Trade Representative for the State of California in London and has since held key leadership positions, including Head of Sales at WIN Plc (now Cisco) and COO at Twistbox Entertainment (now Digital Turbine). Currently, he serves as the CEO of QuickBlox, a leading AI communication platform. Beyond his work experience, Nate is actively involved as an advisor and investor in startups like Whisk.com, Firstday Healthcare, and TechStars. He holds degrees from UC Davis and The London School of Economics and Political Science (LSE).