AI in Healthcare: Turning Automation Into Better Patient Care

Healthcare has spent years asking how AI can help clinicians work faster. Now, as platforms such as Heidi II move beyond clinical documentation to handle chart preparation, referrals and patient follow-ups, a more consequential question is emerging: What if the biggest opportunity isn’t making healthcare more efficient, but changing how patients experience it?

Automating administrative work could reshape the boundaries of a patient visit, the responsibilities of care teams and the way health systems define success. But those gains aren’t guaranteed. Without rethinking the underlying model, organizations risk using sophisticated technology simply to squeeze more appointments into already crowded schedules.

Members of the Senior Executive Healthcare Think Tank, a curated group of experts in patient experience, workforce strategy, policy, quality, equity and healthcare technology, explore what leaders should reconsider as AI takes on more of the work surrounding care—examining the choices that will determine whether automation becomes another productivity tool or a catalyst for more thoughtful, connected and patient-centered care.

“Capacity freed by automation should not become a reason to pack more appointments into the day.”

– Somnath Banerjee, Engineer Lead Senior at a Fortune 50 health insurance company

SHARE IT

Shift From Visit Volume to Value-Based Care

For Somnath Banerjee, Engineer Lead Senior at a Fortune 50 health insurance company, the first change is to rethink how organizations value clinicians’ time. If AI takes over chart preparation, referrals and follow-ups, the resulting capacity should support relationships that extend beyond individual encounters.

“Capacity freed by automation should not become a reason to pack more appointments into the day,” Banerjee says. “Healthcare organizations should use it to rebuild care around continuity.”

He sees this as an opportunity to accelerate the transition to value-based care, where organizations reward outcomes rather than throughput. Nurses and care coordinators should be able to concentrate on the clinical work they are trained to do, while reliable patient data and clear accountability underpin automated processes.

“This is the moment to fully commit to value-based care,” he says. “Every automated action also needs clear accountability, auditability and a human ready to step in.”

For executives, that means connecting automation investments to care models, staffing decisions and reimbursement strategies.

Redesign the Entire Patient Journey

Feri Naseh, Founder and CEO of MeTime Healing LLC, argues that the opportunity extends well beyond individual administrative tasks.

“AI presents healthcare with an opportunity to do more than automate administrative tasks,” Naseh says. “It gives us a chance to fundamentally rethink how care is designed and delivered.”

That redesign starts with the patient’s entire journey, from the first concern through diagnosis, treatment and ongoing support. Instead of allowing organizational boundaries and disconnected workflows to dictate the experience, leaders should strengthen coordination across care teams and use AI to identify opportunities for earlier intervention and personalized support.

“This requires a shift from a volume-driven model of care toward one that prioritizes outcomes, continuity and the patient experience,” she says.

Naseh also emphasizes the workforce implications. Removing repetitive administrative work can reduce cognitive burden and create more room for meaningful patient interactions. But technology alone cannot deliver that benefit.

“The organizations that succeed will be those that pair intelligent technology with accountable clinical leadership, thoughtful workforce redesign and a commitment to treating every patient as an individual—not simply another encounter to process,” she says.

Executives should map the full patient journey before deploying automation, identifying where handoffs fail, where patients become disengaged and where earlier intervention could prevent complications.

Use AI to Make Team-Based Care Work

Jason Foodman, Managing Director at Archetype Growth, believes organizations should use AI to make established models of coordinated care easier to deliver.

“Healthcare organizations know what works in care delivery: team-based, clinician-led, value-based care,” Foodman says. That model coordinates care across providers, places clinical direction with physicians and evaluates success through patient outcomes rather than the number of services delivered.

The challenge has often been operational. Providers must share updates, track referrals and ensure that someone follows through when patients need additional services. Agentic AI can reduce the coordination burden by sharing information across care teams and flagging tasks that require attention.

“Autonomous, agentic AI can reduce that burden by automatically sharing patient updates with care team members, tracking referrals and flagging follow-up tasks that need attention,” he says.

Foodman cautions against treating these capabilities as merely another way to cut costs.

“The promise of autonomous, agentic AI, however, isn’t limited to making administrative processes more efficient and cost-effective,” he says. “AI also enables proven models of care delivery to happen more seamlessly and more effectively.”

“The greatest value of AI in healthcare is the space it can create for human-to-human interaction.”

– Anna Koskinaris, Founder of Supreme Health Solutions

SHARE IT

Protect Time for Human Connection

For Anna Koskinaris, also known as Anna K, Founder of Supreme Health Solutions, the central question is what patients gain when technology reduces the administrative burden on clinicians.

“If AI can take on much of the administrative work, healthcare organizations should use the opportunity to give patients more time, attention and compassionate care,” she says. “Human connection cannot be replaced.”

That requires leaders to reconsider appointment lengths, scheduling expectations and the pressures that encourage clinicians to rush through conversations. Patients bring individual concerns, circumstances and life experiences that may not fit neatly into a template or emerge through a checklist.

“Patients should feel comfortable and confident sharing their concerns and leave knowing they were heard and understood,” Anna says.

A 2019 study published in the Journal of General Internal Medicine found that clinicians interrupted patients after a median of just 11 seconds when inviting them to share their concerns. The findings highlight why giving patients more time to speak—and clinicians more time to listen—can be essential to understanding the full picture of a patient’s health.

“For me, the greatest value of AI in healthcare is the space it can create for human-to-human interaction,” she says. “The time saved should become time invested in people.”

Leaders can translate that principle into practice by protecting time for complex cases, measuring patient-reported experience and training clinicians to listen without prematurely steering the conversation.

Move From Episodic Visits to Preventive Care

Mick Safron, Founder and CEO of Biohackers World, sees AI-enabled administration as an opportunity to make care more proactive.

“If AI absorbs the paperwork around a visit, the win isn’t squeezing in more patients,” Safron says. “It’s changing what a visit is.”

Traditional encounters can offer only a limited snapshot of a person’s health. Meanwhile, wearable devices, continuous glucose monitors and sleep trackers may collect information that never reaches the clinical conversation because reviewing it takes time.

Safron recommends using recovered capacity to support longitudinal care: examining trends rather than isolated readings, considering appropriate preventive testing and helping patients develop personalized approaches to sleep, nutrition, movement and stress management.

“Reinvest the freed hours into longitudinal, preventive care: trend review over single readings, earlier biomarker testing and personalized protocols for sleep, nutrition, movement and stress,” he says.

He also sees an opportunity to strengthen patient engagement by helping people understand their own health information.

“Clinicians can be that interpreter and coach,” Safron says. “Automation only pays off if organizations reinvest time in the relationship, not throughput.”

Measure Completed Care, Not Completed Tasks

Harikrishnan Muthukrishnan, Principal IT Developer at BCBS Florida, argues that leaders need to change who performs the work, what the organization measures and who receives access to services.

“Most of the administrative work that Heidi II handles is done by medical assistants and coordinators, not physicians; rather than cut those roles, reskill them into care navigation, coaching and social-needs screening the system never had a billing code for,” Muthukrishnan says.

That approach makes automation a workforce redesign opportunity rather than simply a cost-cutting exercise. Staff who previously handled repetitive tasks could help patients navigate treatment, overcome barriers to follow-up and connect with community resources.

“Then stop measuring visits and start measuring closed loops: referrals that became appointments, abnormal results acted on in time, discharged patients stable at day seven,” he says.

Finally, Muthukrishnan wants organizations to direct recovered capacity toward patients who struggle to obtain care, including those facing long specialty wait times.

“Automation lets a system serve the people it has been turning away, if leadership chooses to,” he says.

“AI should become the invisible operating layer that removes friction, while care teams focus on empathy, judgment and outcomes.”

– Vikas Gupta, Technical Manager at HCL America

SHARE IT

Make Technology the Invisible Operating Layer

Vikas Gupta, Technical Manager at HCL America, says that healthcare leaders should stop treating AI primarily as an efficiency tool and instead use it to remove friction from care delivery.

“The real opportunity is not to see more patients in less time, but to give clinicians more time for complex decision-making, relationship-building, preventive care and care coordination,” Gupta says.

That distinction separates organizations that automate existing workflows from those that redesign them. A faster administrative process may improve productivity without changing the patient experience. A redesigned process can use the same technology to ensure clinicians have better information, more time for complex cases and stronger support between visits.

“AI should become the invisible operating layer that removes friction, while care teams focus on empathy, judgment and outcomes,” he says.

Gupta’s perspective points to a practical implementation principle: Start with the clinical workflow, not the software’s feature list. Determine where information gets lost, which decisions need human judgment and what clinicians should be able to do differently when repetitive tasks disappear.

“Organizations that simply automate existing workflows will gain incremental value; those that redesign care delivery around human connection and clinical expertise will achieve transformational impact,” Gupta says.

For senior leaders, that means involving clinicians, IT teams and operational managers in the same redesign effort, with shared accountability for the resulting patient and workforce outcomes.

Replace Rushed Visits With Continuous Patient Engagement

Mahendran Chinnaiah, Digital Healthcare Architect at a major U.S. healthcare and pharmacy services firm, believes AI-enabled administration creates an opportunity to move beyond a transactional approach to care and build more continuous relationships with patients.

“Healthcare must abandon the rushed, 15-minute transactional visit model,” Chinnaiah says. “When AI handles chart prep, referrals and administrative coordination, clinical capacity shifts from manual data entry back to longitudinal patient engagement.”

That shift requires more than automating existing workflows. Organizations must rethink how and when care happens, replacing an overreliance on episodic appointments with proactive touchpoints that help care teams respond to patients’ needs over time.

“Instead of clinicians spending half their cognitive load wrestling with screens and referral handoffs, care teams can focus on shared decision-making, complex chronic care management and preventative interventions that actually move clinical outcomes,” Chinnaiah says.

That could mean creating more opportunities for follow-up between appointments, dedicating additional resources to patients with complex chronic conditions and giving clinicians greater capacity to make collaborative treatment decisions.

The goal is to ensure AI does more than accelerate administrative work. By reducing operational friction, organizations can give care teams more room to deliver the continuous, proactive care that patients need.

Build Accountability Into Every Automated Workflow

Jordan Henry, Founder and Chief AI Ethicist at Veritas AI Consulting, argues that the new model of care must preserve continuity while establishing clear ownership of automated work.

“Healthcare organizations should redesign care around continuity, not visit volume,” Henry says. “Use freed capacity for longer conversations, shared decisions, proactive outreach and coordination between visits, rather than simply adding appointments.”

That requires more than deploying agents. Organizations must decide who owns each referral, who follows up when a patient does not respond and which actions require clinician approval. Automation should make responsibilities more visible, not create ambiguity about who is accountable.

Henry points to Heidi II’s clinician oversight and configurable approvals as features worth preserving as automation expands.

“Redesign team roles around patient needs, with clear ownership of referrals and follow-up,” he says.

He also recommends moving beyond technology-centric measures.

“Measure success through outcomes, access, patient trust and clinician workload, not notes completed or minutes saved,” Henry says.

For executives, the practical starting point is an AI governance framework that defines permitted actions, escalation pathways, audit trails and human review requirements. Leaders should test workflows for errors and establish procedures to pause or reverse automated actions when necessary.

Put Diagnostic Accuracy Before Administrative Speed

Dorothy Riviere, CEO and Founder of Work Resilience, raises a critical safety consideration: Automating the work that follows a diagnosis makes the accuracy of that diagnosis even more important.

“If AI automates the referrals and follow-ups, it also automates whatever pathway the diagnosis sets in motion,” Riviere says. “A wrong diagnosis just moves faster.”

Riviere argues that organizations should use recovered time to understand the patient’s history and distinguish the details that matter most.

“So spend the time AI gives back on the patient’s story,” she says. “Studies have found the history alone points to the right diagnosis in about three of four cases.”

That makes careful listening a safety practice, not just a patient-experience initiative. Leaders should preserve time for clinical history-taking, ensure AI-generated recommendations are reviewed appropriately and audit whether automated referrals and follow-ups align with the clinician’s assessment.

The goal is not to slow technology down. It is to ensure that speed serves accurate diagnosis and appropriate care.

Rethinking Care Delivery in the AI Era

• Make continuity the new measure of success. Reinvest time saved through AI in proactive relationships and value-based care rather than filling schedules with more appointments.
• Redesign care around the entire patient journey. Use AI to strengthen coordination, enable earlier intervention and deliver more personalized support.
• Use AI to make team-based care more effective. Automate routine coordination and referral tracking so care teams can focus on delivering better outcomes.
• Protect time for human connection. Rethink appointment schedules and expectations so clinicians can listen carefully and address patients’ individual concerns.
• Turn administrative savings into preventive care. Use recovered capacity to review longitudinal health data and help patients make informed decisions about their health.
• Reskill staff and expand access. Prepare administrative teams for care navigation and coaching while directing recovered capacity toward underserved patients.
• Redesign workflows around clinicians and patients. Treat AI as an operating layer that reduces friction and enables more time for clinical judgment, coordination and relationship-building.
• Replace transactional visits with continuous engagement. Create proactive touchpoints between appointments to support shared decision-making, chronic care management and prevention.
• Build accountability into every automated workflow. Define clear ownership, human oversight and outcome-based measures as AI takes on more administrative tasks.
• Prioritize diagnostic accuracy over speed. Use the time AI saves to understand patients’ histories and ensure automated referrals and follow-ups support appropriate clinical decisions.

Turning Time Saved Into Better Care

The real test of healthcare AI won’t be how many tasks it completes or how many minutes it saves. It will be what organizations choose to do with the time they get back. Leaders can preserve a system built around packed schedules and short encounters—or use this moment to challenge long-standing assumptions about how care should work. That means rethinking staffing, incentives, access and the patient visit itself, with a sharper focus on what people need to get and stay healthy.

For healthcare leaders, the hard part isn’t deciding whether to adopt AI, but deciding what to change once it’s in place. Will clinicians get more time to hear a patient’s concerns? Will someone make sure a referral leads to an appointment? Will patients who have struggled to get care finally get through the door? Those are the questions worth asking—and the answers will matter far more to patients than how quickly the paperwork gets done.