Happy Thursday, Hospitalogists!

Today we're talking about all things agentic AI and primary care as a service - including checking back in with our friend Tom over at Lumeris. Specifically, I'm demystifying and breaking down what primary care as a service is, and what will enable future success. Enjoy! 

Was this email forwarded to you?

SPONSORED BY LUMERIS

Agentic AI embeds tasks into workflows, not to-do lists.

Not all AI is created equal in healthcare. Traditional AI predicts. Agentic AI acts.

Healthcare teams face increasing pressure to improve outcomes and close care gaps, often without the tools or teams to help them do it. Agentic AI is a gamechanger.

Agentic AI integrates directly into workflows, automating tasks, initiating outreach, and freeing clinicians to focus on patient care. This doesn’t just help clinicians take better care of patients. It helps them avoid burnout.

A new white paper from Lumeris reveals the 7 key differences every clinician and health leader needs to know about AI to deliver the best care now.

BLAKE'S BREAKDOWN

Going deeper on an interesting topic, theme, or trend 

Imagine a healthcare system where the bottlenecks of today like six-month waits for a primary care appointment and patients getting lost in follow-up chaos are transformed into seamless, tech-empowered experiences. Imagine having key stressors that lead to physician burnout and patients missing opportunities for preventive care becoming relics of the past. 

That’s exactly the vision the team at Lumeris is championing: a future where Primary Care as a Service™ (PCaaS) becomes the model for how care is delivered, experienced, and scaled, how physicians are supported, and how patient outcomes are improved.

It’s a big vision backed by a team that is writing the playbook for making it a reality, and Tom, the groundbreaking AI-enabled PCaaS platform from Lumeris that is powering the transformation. How does a model like this work?

First, let me remind you a little bit about Tom. You may recall I got my first insider’s look at Tom when it was unveiled at ViVE 2025. I already knew the Lumeris team had been doing great work in value-based care, health system transformation, and primary care for decades. So when they invited me to a special preview, I was excited to see what was next.

The preview didn’t disappoint. The team at Lumeris had been working behind the scenes to combine over 20 years of healthcare research, deep clinical insight, and five years of dedicated AI development to address one of healthcare’s biggest challenges: America’s primary care capacity crisis.

The answer they showed me was Tom, a multi-agent, AI-driven system purpose-built to transform primary care delivery at scale. 

Tom aggregates data to create more complete patient records, connecting billions of clinical and non-clinical data points ranging across multiple sources, including:

• Health systems
• EHRs
• Labs
• Pharmacies
• Claims data
• CMS
• HIE data
• Wearables (including a partnership with Oura)
• Social determinants
• CGMs
• Non-clinical publicly available consumer data

From there, Tom computes the Best Next Action – determined by data-driven algorithms – for the patient within the context of its task. But Tom doesn’t just passively signal a clinician what that action is. Tom takes action like an AI-powered care team member: autonomously scheduling, following up, educating, flagging urgent issues, and closing care gaps behind the scenes.

In fact, Tom acts as a proactive partner inside primary care practices, supporting everything from patient scheduling, to care coordination, to chronic disease management. It’s like care coordination on steroids, delivered by technology.

Needless to say, I was impressed. But what really stood out to me that day, and as I dove deeper into Tom’s capabilities to introduce Hospitalogists to Tom in an earlier newsletter, was how committed Lumeris was to creating and pioneering a new healthcare delivery model: PCaaS.

Inspired by SaaS (Software as a Service), PCaaS represents a paradigm shift in how primary care is delivered. It leverages digital infrastructure, automation, and cutting-edge AI to extend the capabilities of doctors, nurses, and health systems, not replace them. 

This is essential in primary care, where we’re facing staggering resource and capacity issues, including: 

• 87,000 too few primary care physicians by 2037, leaving over 100 million Americans without proper access. 
• The gap between the 500 million available primary care hours today and the 2 billion hours needed to ensure proper access for all Americans. 
• Burnout and administrative overload, which combined with a shrinking workforce, exacerbate the crisis.

Primary care is the entry point to the entire healthcare system. When it falters, the whole system suffers. That’s not a challenge we can solve by adding more humans alone. The sheer scale of the problem demands a new approach. This is where technology, and specifically AI-powered care models, come in. 

AI is a force-multiplier. It doesn't just digitize processes—it redefines capacity. By turning every human into a "super-clinician" supported by intelligent systems, and by enabling machines to safely and autonomously handle low-risk, repetitive tasks, AI-enabled platforms offer the promise of a step-change—not just a small upgrade. Here's how:

AI can process vast data—medical records, lab results, wearable data, and social determinants—to generate clinical insights in real time. This allows physicians to focus on higher-order thinking and patient connection, while AI handles routine triage and decision support.

Agentic AI—autonomous agents capable of carrying out tasks—can perform symptom assessments, medication reviews, and follow-up planning, 24/7. This reduces wait times and extends care reach beyond human availability.

A virtual care assistant can take on time-consuming but critical tasks like medication reminders, post-discharge followups, appointment scheduling, and answering patients’ questions. This lifts the administrative burden from doctors and care teams, freeing them to focus on what only humans can do. And by reclaiming clinicians’ time, we ease burnout and indirectly expand system capacity.

AI-driven platforms can analyze patient risk profiles and nudge them toward screenings, vaccinations, and checkups—using personalized, behaviorally informed outreach that integrates with mobile apps or messaging platforms.

Point solutions that solve just one problem, like post-hospital discharge calls, have dominated digital health for years. All well-intentioned, but let’s face it: healthcare has a lot of problems to solve. If you start implementing one tool for intake, another for scheduling, another for analytics, and so on, the fragmentation can create more inefficiencies, not fewer. And one-off fixes don’t talk to each other. Today it’s like we’re in the middle of our own science fiction movie, with humans talking to AI and AI tools talking to each other. 

What’s required now is a unified operating system for primary care—one that blends:

• Data interoperability across EMRs, labs, and patient devices.
• Real-time clinical guidelines embedded into care delivery.
• Agentic AI for autonomous task handling.
• Contextual insights for human and AI-assisted decision making.
• Admin-saving automations for documentation, billing, follow-ups.

This "platform-first" model reduces dependency on manual labor, while expanding care availability and consistency at scale. It’s the only viable path to closing the primary care gap and solving for both quality and access. 

When I look at a solution like Tom, it’s the kind of end-to-solution we’re talking about. Tom, for example, is designed to integrate across systems, connect with both big and smaller EHRs, scale across nearly 100 payer data streams, and leverage data with real-time clinical guidelines, all underpinned by agentic AI. This isn’t patchwork; it’s architecture.

Before I give my final thoughts on PCaaS, I wanted to share some of the changes that have happened since I last saw Tom.

To bring their ambitious vision of PCaaS to life, Lumeris has teamed up with some of the most influential leaders in tech and science. They’ve forged three powerhouse partnerships, each contributing unique expertise to the platform’s evolution.

Lumeris partnered with Google Cloud to supercharge Tom with next-gen AI and cloud secure cloud infrastructure. 

Google Cloud’s powerful infrastructure and models, including Gemini Flash and Chirp3-HD, help transform Tom into an intelligent, conversational assistant embedded directly into clinical workflows.

The collaboration enables Tom to handle broad-based population health level outcomes, turning fragmented data into actionable insights, improving care coordination, and delivering deeply personalized patient experiences at scale.

As Thomas Kurian, Google Cloud’s CEO, put it: “By combining Google Cloud’s AI capabilities with Lumeris’ primary care insight, we can provide doctors and nurses with the support they need to improve care for their patients.”

To ensure Tom delivers clinically sound and evidence-based recommendations, Lumeris integrated Wolters Kluwer Health’s UpToDate®, the current gold standard in clinical decision support.

This integration equips physicians with real-time guidance and recommendations at the point of care, helping them make faster, more confident decisions without added complexity. Tom synthesizes EHR and claims data into a 360-degree view of each patient, while UpToDate layers on the latest medical evidence — all seamlessly embedded in physicians’ workflows.

For busy care teams, this means Tom can suggest and even act on best next steps, like scheduling screenings, managing medications, or escalating concerns, improving outcomes at scale. And by combining Tom’s AI insights with UpToDate’s trusted guidance, clinicians can act earlier and improve outcomes.

The partnership with the MIT Computational Biology Lab, led by Professor Manolis Kellis at MIT CSAIL, takes Tom’s capabilities to the next frontier: precision medicine and predictive modeling.

By integrating advanced computational techniques, including dynamic patient modeling and trajectory analysis, Tom will not just react — it will anticipate future health events and guide interventions proactively. This work includes understanding environmental, socioeconomic, and molecular factors, allowing the platform to tailor care especially for underserved populations.

As Professor Kellis describes, the collaboration is about building “AI models that go beyond predicting outcomes, seeking to actively shape care by optimizing interventions that improve long-term health.”

Together, these collaborations create an ecosystem where AI doesn’t replace doctors — it amplifies them. Google Cloud brings scale and power; UpToDate ensures medical accuracy; MIT pushes the envelope on personalization and predictive insights.

Primary care is under strain like never before. PCaaS, driven by AI, data, and cutting-edge platform thinking, has the potential to fundamentally transform how we deliver primary care at scale and keep populations healthy. But it requires bold investment and a mindset shift...from solving surface-level problems with disconnected tools, to redesigning the system itself.

PCaaS gives us a real shot at impact. The category is still being refined but it’s clear that PCaaS is no longer just an internal idea at Lumeris. It’s becoming part of the healthcare conversation. Media mentions are rising, and while competitors are nibbling at the edges, Lumeris is one to watch.

SPONSORED BY LUMERIS

MISCELLANEOUS MADDENINGS

That's all for this week! Enjoy the weekend Hospitalogists. Next week will be a doozy with most companies having announced earnings.

Thanks for the read! Let me know what you thought by replying back to this email.

— Blake  

Get your brand in front of 46,500+ executives and healthcare decision-makers.

I'm building a community of leaders in strategy, finance, and ops
at hospitals and health systems to help us connect, learn, and grow together.

Join the movement here.

Workweek Media Inc. 
1023 Springdale Road, STE 9E
Austin, TX 78721

Want to ruin my day? Unsubscribe.