Agentic AI Empowers Doctors, Streamlines Healthcare Administration

Forbes

30-05-2025

Medical students today increasingly face questions arising from the rise of AI. How will AI transform the nature of the profession? Will artificial intelligence make some medical roles obsolete, or will it enhance medical professionals' ability to care for their patients? Abdel Mahmoud, a British-trained physician and CEO of Anterior, firmly believes in the latter. Mahmoud, whose background includes roles at Facebook and Google, is leveraging his unique blend of clinical expertise and technological insight to reshape the healthcare administrative landscape through agentic AI.
Anterior AI is an innovative platform developed by clinicians specifically for clinicians, aiming to simplify cumbersome administrative workflows that bog down healthcare systems. Typically, doctors spend over two hours a day on administrative deskwork. At the heart of Anterior AI's approach is "Florence," an AI assistant named in honor of the founder of modern nursing, Florence Nightingale. Florence's primary role is to streamline interactions between healthcare providers and insurance companies, particularly regarding prior authorizations, which are traditionally one of the most time-consuming and frustrating processes in medicine.
"Hundreds of thousands of nurses, from the U.S. to the Philippines, spend their days buried in faxed medical records, checking boxes rather than caring for patients," Mahmoud explains. By automating much of this tedious work, Florence significantly accelerates decision-making, reducing the review time from hours or days to minutes.
Florence illustrates the potential of the agentic AI approach when confronting the bureaucratic morass. During the live demonstration, Mahmoud showcased how Florence quickly summarized a patient's complex medical history from multiple faxed documents, clearly highlighting relevant clinical details associated with the insurance provider's criteria. Florence then identified a gap in documentation—specifically, proof of supervised physical therapy—and independently initiated a phone call to the healthcare provider. After acquiring the missing information through a brief conversation, Florence promptly updated her analysis and confirmed the authorization. This process took only minutes, demonstrating how swiftly Florence can facilitate complex administrative decisions. Also noteworthy was the natural, human-centric approach to the user interface, with the AI agent interacting in a warm conversational tone and doing so in a regional accent selected to suit the location being called.
A noteworthy aspect of Anterior's approach is its commitment to responsible deployment. Florence currently handles only prior authorization approvals but does not pursue denials. "We intentionally don't automate denials," Mahmoud emphasizes. 'Our goal is to rapidly provide certainty for the 95% of treatments that inevitably get approved anyway, not to create new barriers." This should help insulate Florence and the rest of Anterior from potential threats by agentic anti-insurance agitators.
This human-centered AI philosophy runs deep in Anterior's operations. Anterior engages healthcare professionals as clinician scientists, utilizing nurses and doctors to develop and refine sophisticated AI workflows directly. Rather than reducing opportunities, Mahmoud argues that tools like Florence should actually broaden career paths within medicine. "This wave of AI will allow healthcare professionals to finally do what they're trained for, namely to care for people, not PDFs," Mahmoud predicts.
Currently, Anterior is collaborating with several major entities involved in healthcare administration and expects to handle approximately four million prior authorization decisions by year's end. The initial feedback has been overwhelmingly positive, reducing delays, lowering operational costs, and significantly improving job satisfaction among both clinicians and administrators.
Looking ahead, Mahmoud envisions an evolving medical education system where students learn early to interact seamlessly with AI. He advises future clinicians to view themselves increasingly as "editors and coaches," critically guiding AI systems rather than merely serving as passive users of the technology. "While AI won't replace medical professionals, it will redefine their roles," Mahmoud assures. "The clinicians of tomorrow must be trained not only in patient care but in managing, auditing, and refining the AI-driven tools that will inevitably underpin healthcare workflows."
Mahmoud's advice to medical students and young doctors is clear: embrace this new era, understand how these agentic AI systems function, and prepare to be actively involved in shaping their ethical and effective use. By doing so, they can significantly improve patient outcomes, make administrative burdens almost invisible, and restore the essential human connections at the heart of medicine. "The future of healthcare won't just be more efficient—it will be fundamentally more humane," Mahmoud concludes.