Swaay Health's Podcast of the Year
X

Right now, uncertainty is the only constant in healthcare. Leaders are grappling with budget cuts, shifting policies, legal challenges to long-standing mandates, and an exhausted workforce. The landscape is evolving rapidly, often without clear guidance on what’s next or how to prepare.

But throughout this unpredictability, one thing hasn’t changed: people still get sick. Families still need answers. Communities still rely on timely, high-quality care. In fact, the need for accessible, efficient healthcare has never been more urgent.

This is not a moment to pause. It’s a moment to refocus — and double down on patient access and operational efficiency.

The New Normal of Uncertainty

Across the industry, leaders are feeling the squeeze from every angle. At the policy level, questions around Medicaid expansion, telehealth reimbursement, and DEI initiatives have created a fog of confusion. Funding windows open and close without warning. Priorities shift seemingly overnight. And no one’s quite sure what will be funded — or when.

Operationally, the pressure is relentless. Staffing shortages are hitting hard, not just in clinical roles, but also in IT departments and access centers. Burnout is no longer just a concern; it’s the reality. At a recent industry event, one executive remarked that call volumes at their access center were spiking — not because of new demand, but because frustrated patients couldn’t navigate existing digital tools. The system is overloaded.

Financially, most health systems are operating with little room for error. Margins are flat or shrinking. IT teams are being asked to stretch aging systems further while driving innovation on tighter budgets. In this environment, every inefficiency becomes a liability.

And then there’s the patient experience — the part of the story that can get overshadowed. 

This isn’t just noise. It’s a call to act.

The debates about policy are ongoing, but one thing is clear: people still need care. Delaying that care doesn’t make the need go away, it just makes it more urgent later.

What Happens If We Wait

In times of uncertainty, it’s tempting to hold still. To wait for clarity before making changes. But in healthcare, waiting often makes things worse.

Delaying or deprioritizing access initiatives won’t stabilize the system — it will destabilize it further:

And once patients disengage from the system, rebuilding that trust is possible but it takes more time, and money, to bring them back. 

Focus, Not Freeze

Instead of freezing, healthcare organizations must focus. That means getting smarter about where and how they invest in access and efficiency.

Lean Into Efficiency

Efficiency doesn’t mean doing more with less. It means doing the right things, better.

Small improvements in how appointments are booked, how reminders are sent, or how patients are guided through the system can lead to big wins — for both experience and revenue.

Prioritize Access

Even modest improvements in scheduling or communication can yield major results:

Remember: prioritizing access is essential to healthcare because access is the gateway to outcomes. If patients can’t get in the door (whether that door is physical, digital, or operational) nothing else in the care journey can happen.

Build Resilience Into Your Tech Stack

Technology, too, plays a key role. The future isn’t about ripping out what’s already in place. It’s about building resilience into what’s already working. 

This isn’t the time for short-term patches. Invest in systems designed to evolve:

When done right, these investments don’t just help organizations weather uncertainty. They make them more agile, more adaptive, and ultimately more effective.

Leading Through Uncertainty: A Call to Action

This moment calls for leadership — not paralysis. The health systems and clinics that come out stronger won’t be the ones that waited. They’ll be the ones that acted with purpose, even amid ambiguity, in order to:

Because at the end of the day, uncertainty doesn’t change the mission. It clarifies it. If anything, uncertainty makes it clearer than ever: every patient deserves access to timely, compassionate, and efficient care. People still need care. They always will. And the systems that serve them need to be ready — not someday, but now.

In healthcare, solving challenges always requires more than a single step. From scheduling appointments and verifying insurance to managing care transitions and reducing no-shows, the path to better patient outcomes is rarely straightforward. Just as healthcare organizations rely on interconnected systems like electronic health records (EHRs) and practice management software, AI solutions are most effective when they work together seamlessly.

This is where agentic AI comes in. Agentic AI refers to a system of specialized AI agents—each designed to perform specific tasks—working in coordination to achieve a larger goal. Think of it as a team of experts, each contributing their skills to move a patient smoothly through their care journey. By handling tasks autonomously, these agents reduce administrative burdens and allow healthcare teams to focus on what they do best: caring for patients.

At Luma Health, we believe the future of healthcare lies in the collaboration of AI agents. Our AI-powered Navigator solution uses agentic AI to streamline workflows, improve patient experiences, and drive measurable outcomes. From automating routine tasks to providing actionable insights, our network of AI agents supports both operational and patient care teams in delivering exceptional care.

In this blog post, we’ll explore how multiple AI agents work together to solve complex healthcare challenges, the benefits of agentic AI, and how Luma Health is helping providers navigate this new era of intelligent automation.

What is Agentic AI and How Does it Work?

Agentic AI is like having a team of digital assistants, each with its own role, working together to achieve a common goal. Each agent is specialized, meaning it has a clear task—whether that’s gathering data, analyzing information, or triggering actions.

As Ivan Viragine, AI Engineering Manager at Luma Health, explains: “An agent is a combination of a large language model (LLM), a prompt, and a set of tools. In Navigator’s case, we have one agent for verifying a patient’s identity, another for listing appointments, and others for tasks like confirming visits. These agents work together to achieve their goal of understanding and fulfilling the user’s request.”

These agents coordinate in real time, adjusting their actions based on new information. For example, if a patient cancels an appointment, one agent verifies the patient’s identity, another lists the upcoming appointments to confirm which one to cancel, and a third cancels it directly in the EHR. This intelligent division of labor reduces administrative burden and ensures patients receive timely care.

This multi-agent or “agentic AI” approach also improves accuracy and reliability. Instead of relying on a single AI to parse an overwhelming set of rules or data—like expecting one person to memorize and apply a 100-page manual—agentic AI distributes the cognitive load. Each agent focuses on a smaller, well-defined domain (like one chapter of that manual), and a coordinating supervisor directs requests to the most relevant agent. This specialization not only speeds up performance but also reduces the risk of error or confusion—especially critical in healthcare, where mistakes can have serious consequences. The result is a system that’s more capable, precise, and less prone to “I can’t help you with that” dead-ends.

Why Healthcare Needs Agentic AI Systems

Healthcare operations are inherently complex. From scheduling and follow-ups to prior authorizations and patient communications, these processes often require complex coordination with large groups of people. Disconnected systems lead to inefficiencies, delays, and frustrated patients. On top of that, healthcare staff are burdened with administrative tasks—research shows that clinicians spend nearly 50% of their time on paperwork and administrative work, taking away from patient care.

Healthcare needs multi-agent AI systems. Many real-world patient interactions are too nuanced for a single agent to manage. Imagine a patient who wants to cancel her upcoming PCP appointment and refill her child’s prescription—all within one call to the clinic’s access center. A common scenario, yet far too intricate for traditional, monolithic AI systems to handle effectively. This is where multi-agent AI shines.

One way to understand the power of agentic AI is through analogy: imagine asking a person to memorize a 100-page instruction manual and then locate the answer to a very specific question. The likelihood of them missing or mismanaging the task is high—because the answer might live in a tiny paragraph on page 56. But if you break the manual into chapters and assign a smaller expert to each one, then have a supervisor route questions to the right expert, accuracy improves dramatically. Each specialized agent only needs to sift through a narrow slice of information, which significantly reduces the risk of misunderstanding or error.

This is especially critical in healthcare, where the consequences of a mismanaged task can directly affect patient safety or delay care. Smaller, specialized agents reduce cognitive load and hallucinations—two well-documented risks in large language models—resulting in more reliable performance.

Think of agentic AI like a surgical team. Each member—surgeon, anesthesiologist, nurse—has a well-defined role. Similarly, AI agents specialize in distinct functions:

As Hwee Min Loh, Senior Product Manager at Luma Health, describes it: “Agentic AI means that the reasoning framework is spread across multiple specialized agents, rather than relying on one massive list of instructions. This reduces common issues like hallucination and enables more accurate, reliable outcomes. Navigator ensures all patient requests are directed to the right specialized agents.”

Real-World Use Cases of Agentic AI in Healthcare

At Luma Health, we see the impact of agentic AI systems every day. One compelling example is our work with UAMS (University of Arkansas for Medical Sciences). Faced with rising call volumes and patient communication challenges, UAMS partnered with Luma Health to deploy our Navigator AI platform.

With Navigator’s agentic AI approach:

The results were transformative—UAMS saw a 20% decrease in patient no-shows and significantly reduced call center volume. Staff were freed from repetitive tasks, allowing them to focus on providing high-value care. Learn more in our UAMS case study.

Looking Ahead — The Future of Agentic AI in Healthcare

The future of multi-agent AI in healthcare is exciting. As AI systems become more adaptive, predictive, and personalized, hospitals will increasingly rely on agentic AI to anticipate patient needs and proactively manage care.

Healthcare organizations will further integrate agentic AI for proactive care management—reducing administrative burdens, improving operational efficiency, and ultimately enhancing patient outcomes. At Luma Health, we’re excited to continue leading this transformation, empowering providers to deliver exceptional care through the power of many AI agents.

AI Agents Must Collaborate

No single AI agent can solve healthcare challenges alone. Just as healthcare providers work as teams, AI agents are most effective when they collaborate to streamline operations and enhance patient care.

At Luma Health, we are committed to applying the power of multi-agent AI to make healthcare easier for providers and patients alike. By leveraging Navigator’s agentic workflow, we help healthcare organizations reduce administrative burdens, improve operational efficiency, and ensure patients receive timely, high-quality care.

Want to see how multi-agent AI can transform your organization? Learn more about Navigator or request a demo today to experience the Luma Health difference.

Frequently Asked Questions

How does agentic AI integrate with our existing systems like EHRs or practice management software?

Luma’s Navigator platform is designed to integrate seamlessly with leading EHRs, scheduling systems, and communication tools. Whether you’re using Epic, Cerner, Athenahealth, or other platforms, our AI agents can access and act on real-time data through secure API connections and industry-standard integrations.

What is the implementation timeline? How much lift is required from our IT team?

Most organizations see their first AI workflows live within weeks. Our dedicated team handles the heavy lifting of that first implementation, with minimal demands on your IT resources. We tailor the rollout to your existing workflows and provide hands-on support to ensure a smooth transition. There are also self-serve tools available so that you can build your own workflows in a no-code, easy-to-use interface using Navigator’s individual agentic AI skills. 

Is this secure and HIPAA-compliant?

Yes. Security and compliance are non-negotiable. Luma Health is HIPAA-compliant and HITRUST-certified, and all agent actions are fully auditable. Our platform ensures patient data is handled securely at every step, with encryption and strict access controls built in.

Will Navigator replace my staff?

Navigator is designed to augment, not replace, your team. By taking on repetitive and time-consuming tasks like scheduling, appointment reminders, and eligibility checks, our AI agents free up your staff to focus on high-value interactions that improve the patient experience and operational outcomes.

What happens when the AI doesn’t know what to do?

Agentic AI is built to collaborate—with each other and with your team. When an agent encounters a complex or ambiguous request, it automatically escalates the task to a human staff member. Patients never hit a dead end, and your team is always in the loop.

How do I measure the ROI of agentic AI?

Navigator includes built-in analytics that track key performance metrics—reduction in call volume, no-show rates, scheduling efficiency, and more. Our clients often see measurable impact within the first few weeks of deployment.

Can we customize workflows or agent behavior?

Yes. Each Navigator agent can be configured to fit your needs, such as modifying the welcome and end message, the agent’s voice, and the action it can take with your patient. Whether you want agents to follow specific scripts, recognize custom intents, or trigger internal protocols, we give you the flexibility to stay in control. There are also self-serve tools available so that you can build your own workflows in a no-code, easy-to-use interface using Navigator’s individual agentic AI skills. 

How do patients feel about interacting with AI agents?

Patients appreciate fast, 24/7 access to the help they need—without waiting on hold. Our Spark AI is designed to be transparent and patient-friendly, clearly indicating when they’re interacting with a digital assistant. When needed, agents seamlessly hand off to staff, ensuring a smooth and trusted experience.

Changing your electronic health record (EHR) system is one of the most significant technology decisions a healthcare organization can make. But while most teams are laser-focused on the EHR transition itself, they may miss a critical opportunity: reevaluating their broader tech stack.

At Luma, we’ve seen firsthand how organizations like University Hospitals took a comprehensive approach to their EHR migration. Stacy Porter, who previously served as the VP of Digital Transformation at University Hospitals in Cleveland, says: “When you’re implementing a new EHR, it’s not just an opportunity — it’s an obligation to look at your entire digital portfolio.”

Here’s why evaluating your IT landscape during an EHR transition can unlock long-term value and how Luma can be an invaluable partner in the process.

1. Consolidate and Simplify Your Digital Portfolio

An EHR change often reveals redundancies and inefficiencies. Before University Hospitals switched to Epic, according to Porter, they completed a capability mapping exercise to compare their existing digital tools — including platforms like Salesforce, Conversa, and RevSpring — against Epic’s capabilities. This allowed them to make informed decisions about what to keep, what to replace, and where gaps existed.

Rather than defaulting to piecemeal solutions like Twilio for patient reminder texts, Porter asked: “Is there a vendor that can consolidate these functions and provide additional value?” Enter Luma Health. By selecting a partner like Luma, they reduced their vendor sprawl and ensured seamless interoperability with Epic from day one.

“When you’re implementing a new EHR, it’s not just an opportunity — it’s an obligation to look at your entire digital portfolio.”

Stacy Porter, former VP of Digital Transformation at University Hospitals

2. Co-Design for Long-Term Success

When evaluating vendors during an EHR transition, consider how co-designing can prevent future headaches. Porter emphasized this strategic approach: “We co-designed with Luma, so when we turned on both Epic and Luma, everything worked by design — no overlap, no retrofit.”

By collaborating with Luma early in the process, University Hospitals avoided unnecessary IT buildout later. This streamlined implementation and reduced the burden on their IT team, with Luma handling most configurations and only needing operational input.

3. Minimize Change Fatigue

For both patients and staff, transitioning to a new EHR means significant change. University Hospitals took a “rip off the Band-Aid” approach to minimize disruption. “Change once, change deep,” Porter said. Rather than subjecting patients and staff to waves of adjustments, they implemented Epic and Luma Health simultaneously.

This reduced the need for multiple rounds of training and communications, ultimately leading to smoother adoption and fewer frustrations.

4. Communicate Effectively

One of the most critical factors in a successful transition is proactive communication. Porter highlighted the importance of clear, consistent messaging to both patients, staff, and providers about what’s changing and why. Now imagine repeating that process three, six, or nine months later when introducing another solution — it’s a scenario best avoided.

By implementing Luma alongside Epic, University Hospitals ensured that everyone was aligned and informed upfront, reducing confusion and frustration.

5. Choose the Right Partners

Not all vendors are equipped to navigate the complexities of an EHR transition. Aditya Bansod, Luma’s co-founder, advises organizations to think holistically: “Every EHR conversation is part of a larger IT conversation. Use this as a moment to clean up your tech stack. Luma can be part of that.”

By choosing a partner like Luma, healthcare organizations can consolidate disparate tools, bring legacy systems into the future, and maximize their EHR investment.

EHR conversions take a lot of consideration from all fronts, and these five reasons make it clear: an EHR migration isn’t just a system switch — it’s a strategic opportunity to modernize and streamline your entire digital infrastructure. Here’s a quick recap to guide your planning:

An EHR migration is not just a software upgrade — it’s a pivotal moment to evaluate and optimize your entire digital strategy. With the right planning, stakeholder engagement, and vendor support, your organization can turn this period of change into a long-term advantage.

At Luma Health, we’re here to help you make the most of your EHR investment. Let’s reimagine what’s possible together.

At Luma, we believe the healthcare industry has moved beyond the initial hype of artificial intelligence (AI). While AI once dominated conversations with grand promises and speculative claims, it has now become table stakes. The industry is entering a new phase, one where AI is treated like any other technology investment — evaluated with rigor and held accountable for delivering real outcomes. For healthcare leaders, this shift requires a focus on practical applications and measurable impact.

AI should not be designed or deployed with the intent to replace your workforce. Instead, it enables your workforce to focus on patient-facing interactions over administrative busywork.

To explore what this shift looks like in practice, in this blog post we’ll follow the journey of Hayes Valley Health Center, a mid-sized hospital navigating the realities of AI adoption. Hayes Valley is fictional, but its challenges are anything but. Modeled after the experiences of Luma customers, the health center faces mounting pressure to improve operational efficiency, deliver high-quality care, and enhance experiences for both patients and staff. Like many health systems, they’re excited by AI’s potential but have struggled to move beyond the buzzwords.

Throughout this post, we’ll explore how Hayes Valley Health Center approaches AI adoption with a focus on outcomes, applying the same level of diligence and strategic thinking they would with any other technology. From identifying clear goals to measuring success, their story serves as a practical guide for healthcare organizations striving to turn AI’s promise into real-world progress.

AI as a Productivity Booster, Not a Workforce Replacement

Like many health systems, Hayes Valley’s leadership initially viewed AI as a way to reduce staff costs. They hoped that chatbots could replace human schedulers, handling appointment management and patient inquiries with minimal human intervention. It seemed like a quick win to drive efficiency and lower expenses.

The reality didn’t match expectations. While the AI chatbot excelled at simple tasks, it struggled with more complex patient requests. Patients seeking specialized care, needing to reschedule complicated procedures, or asking detailed insurance questions grew frustrated. Staff often had to step in without sufficient context, leading to inefficiencies and dissatisfaction on both sides.

Recognizing this challenge, Hayes Valley recalibrated their approach. Instead of positioning AI as a replacement for human schedulers, they deployed it as a productivity booster. The AI was assigned three key tasks:

This shift freed staff to focus on higher-value patient interactions. Without the burden of routine tasks, they could provide more compassionate and personalized support. AI operated behind the scenes to streamline workflows, while human schedulers brought empathy and expertise to complex situations. The result was a better experience for both patients and employees.

Luma’s AI-powered Navigator product supports this kind of balanced approach. Navigator uses conversational AI to assist with appointment management and patient inquiries, ensuring staff can dedicate their time to what matters most. By handling the repetitive, AI enables healthcare organizations to scale their services without sacrificing quality.

The key lesson Hayes Valley learned is clear: AI is not a substitute for human expertise. Instead, it’s a powerful tool to enhance productivity, reduce operational friction, and improve care experiences. When thoughtfully applied, AI empowers staff to excel in their roles, making healthcare more efficient and empathetic for all.

Prioritizing Real Use Cases Over Shiny Tech

What can all of us learn from the missteps of the fictional Hayes Valley? We must shift focus to AI solutions designed to solve practical, high-impact problems, and it should integrate seamlessly with existing systems and staff workflows.

Before adopting any new AI tool we recommend asking four critical questions, based on input from health system leaders we work with who have successfully adopted and deployed AI:

Future-proofed, AI-native platforms like Luma’s are designed with these principles in mind. Built to adapt and grow with healthcare organizations, they ensure today’s AI solutions won’t become tomorrow’s technical debt. Platforms that anticipate industry needs and prioritize interoperability are the ones that will drive lasting impact.

Moving Forward with Practical AI

Hayes Valley Health Center’s journey offers lessons for effectiveAI adoption — avoid the hype, invest in practical solutions, and ensure AI tools support (rather than replace) healthcare teams.

By aligning AI investments with clear goals, empowering staff with adaptable tools, and maintaining a disciplined roadmap, organizations can drive meaningful improvements.

The key takeaway is simple: AI should empower, not complicate. With the right mindset and technologies, AI can become a powerful tool for transforming healthcare—enhancing experiences, improving outcomes, and making care more accessible for all.