New Delhi: As health systems worldwide grapple with workforce shortages, rising chronic disease burdens, and widening inequities, Infosys is positioning “agentic AI” as a key lever to help close the healthcare divide.
In an exclusive interview with ETHealthworld, Venky Ananth, Executive Vice President and Global Head of Healthcare at Infosys, described agentic AI as a shift from passive analytics to proactive action—an evolution he believes is critical for improving access to care at scale.
For a non-technical audience, agentic AI refers to systems that can act autonomously within defined boundaries. They gather information, complete tasks, and coordinate steps on behalf of humans.
Unlike traditional AI models that primarily analyze data and generate insights, agentic AI systems can initiate follow-ups, schedule appointments, remind patients about medications, and escalate cases to clinicians when needed. This distinction, Ananth said, is particularly relevant in healthcare, where access is often limited by “time, distance, and administrative friction.”
“Agentic AI acts as a force multiplier for overstretched systems,” he noted. “It enables care to reach communities that struggle with availability, affordability, or geographic isolation.”
Targeting the Most Urgent Gaps
Globally, Infosys sees three urgent gaps: uneven access to primary care, acute workforce shortages, and persistent disparities in outcomes among vulnerable populations.
To address these, the company has embedded agentic workflows into platforms such as Infosys Topaz Fabric for rapid AI agent development, Infosys Helix for core administration, Wingspan for workforce upskilling, and its Responsible AI toolkit for governance and safety.
“These initiatives are designed to expand reach, simplify navigation, and support early intervention,” Ananth said. “We prioritize solutions that help patients find care faster and reduce delays caused by administrative complexity.”
By focusing on operational bottlenecks prior authorizations, appointment scheduling, and follow-up tracking the company aims to tackle barriers that disproportionately affect underserved groups.
Bridging Gaps in Underserved Communities
In resource-constrained settings, coordination gaps can be as damaging as infrastructure gaps. Ananth believes agentic AI offers a distinct advantage here.
“Traditional AI analyzes data. Agentic AI acts on it,” he said. “That capability matters where resources are scarce and tasks fall through the cracks.”
For example, AI agents can identify rising-risk patients, initiate preventive outreach in local languages, guide them through digital triage, and ensure follow-ups are completed. Crucially, these systems operate under clinician-defined rules and escalate cases when human oversight is required.
“This combination of initiative and supervision makes agentic AI more capable of bridging access gaps than passive predictive models alone,” he added.
Among the most promising applications are digital triage, population health outreach, automation of prior authorizations, and chronic disease management. Health systems deploying agentic workflows are reporting faster care navigation and more consistent patient engagement, particularly in rural and low-income communities.
“These solutions help shift from reactive care to proactive engagement,” Ananth said. “Access becomes not only faster, but more equitable.”
Supporting, Not Replacing, Clinicians
With global shortages of doctors and nurses, concerns about AI replacing human judgment persist. Ananth was clear:
“AI agents support, not replace, clinical decision-making.”
According to him, agents can summarize patient information, gather evidence, and suggest next steps, but recommendations remain fully reviewable. Infosys embeds “human-in-the-loop” controls to ensure clinicians validate all actions.
“AI also flags uncertainties instead of guessing,” he said. “The principle is simple: clinicians decide, AI assists.”
By automating administrative tasks, he added, clinicians gain more time for complex decision-making potentially improving patient safety.
Adoption, however, depends on workforce readiness.
“Simulation-based practice helps clinicians integrate AI into daily routines confidently,” Ananth said. “The goal is for AI to feel like a reliable teammate, not a disruptive intrusion.”
Safeguards Against Bias
As healthcare AI expands, concerns about algorithmic bias remain front and center. Infosys incorporates bias-detection frameworks, diverse training datasets, and continuous performance monitoring into its systems.
“Design principles include inclusive personas, multilingual accessibility, and careful evaluation of social determinants,” Ananth said.
Agentic systems are designed to escalate ambiguous cases for human review rather than making uncertain decisions.
On the question of accountability, Ananth was unequivocal:
“Accountability must always remain with the healthcare organization and the licensed professionals overseeing care.”
Infosys advocates transparent audit trails, clear escalation pathways, and governance structures that ensure all AI recommendations are reviewable.
“Technology can assist,” he said, “but clinical and organizational accountability cannot be delegated to AI.”
From Episodic to Continuous Care
Looking five years ahead, Ananth believes agentic AI will be most transformative in primary care access, chronic disease management, and home-based care.
By integrating multimodal data from electronic health records, wearables, imaging, laboratory reports, and social determinants, AI agents could help shift health systems from episodic treatment to continuous, proactive support.
“Our goal is to help deliver high-quality healthcare to every community, regardless of geography or income,” Ananth said.
If successful, agentic AI may not just optimize healthcare operations—it could redefine how access itself is delivered, making care more anticipatory, inclusive, and equitable in the years ahead.
