Commure was founded within General Catalyst by Hemant Taneja and a core team, working closely with Thomas Jefferson University and HCA. The core thesis behind Commure was simple: if you pair top Silicon Valley technologists with leading health system operators and hospital leaders, you can build better healthcare technology than what exists today.

At the heart of Commure is CommureOS, a data integration engine designed to connect the many disparate software solutions used in healthcare. These disconnected systems place a significant burden on IT teams and often act as a bottleneck to adopting new technology. CommureOS helps solve this by making integration seamless, reducing the workload on IT teams and enabling faster technology adoption within health systems.

A useful comparison is Palantir Foundry—its ontology system makes it easy to build applications on top of existing data. Similarly, CommureOS allows us to quickly integrate with health system data, leveraging pre-built embeddings, workflows, and automation tools. This makes deploying custom applications for a given health system dramatically easier.

When Commure started six or seven years ago, we likely didn’t anticipate the full scope of what we’ve built today—a comprehensive suite of applications ranging from clinical documentation to back-office automation and full-stack revenue cycle management. But when the core of your engineering and product team is focused on seamless integration, it naturally enables rapid expansion and innovation.

If you look at any healthcare organization today—from large health systems down to small private practices—one of the biggest challenges is work tax. For every high-value interaction between a physician or nurse and a patient, there’s an enormous administrative burden: documentation, paperwork, insurance processing, and compliance tasks. This inefficiency creates a drag on the entire American economy.

At Commure, we want to change that. Our vision is to build applications, infrastructure, and automation that reduce this burden—shifting the current 1:10 ratio of clinical work to administrative overhead to 1:5, and eventually 1:1. Achieving that could return $1–$2 trillion in non-inflationary stimulus to the U.S. economy, simply by making healthcare more efficient. That’s money that goes back into the middle class—driving innovation and economic growth.

Beyond the economic impact, automation is also a key lever for improving healthcare outcomes and lowering costs. The high price of American healthcare isn’t due to greedy clinicians or corrupt institutions—it’s driven by excessive bureaucracy and administrative complexity. If we can eliminate that inefficiency, the U.S. could deliver the best healthcare in the world at the lowest cost—just as we do in many other industries.

Our vision spans three core pillars—Clinicians, the Health System, and Patients. On the clinician side, our flagship offering is our Commure Ambient Suite, which encompasses three tiers of ambient documentation solutions: Ambient AI, Ambient Assist, and Ambient Live. These tiers map to different clinical workflows, from fully automated note generation to hybrid AI-human scribing. For example, if a physician wants a purely AI-driven solution on a lighter day, they can use Ambient AI. If they’re under heavy load or dealing with more complex cases, they can flip a switch to Ambient Assist or Ambient Live to involve a medical documentation specialist.

Beyond documentation, we also incorporate care nudges, autonomous coding, and an AI copilot that offers context-aware responses to physician questions. For healthcare administrators, we deliver full-stack Revenue Cycle Management (RCM) solutions, asset tracking (using the same RTLS tech that powers our duress alerting system Strongline), and automation for back-office workflows. And on the patient side, we provide robust patient engagement tools, including pre- and post-visit care plan management—thanks to our acquisitions of RxHealth and Memora. All of these capabilities run on CommureOS, our data integration layer that ties everything together.

No two health systems follow the exact same journey with Commure because our product suite is expansive, covering everything from real-time location services and staff safety tools (like Strongline) to revenue cycle management and AI-powered documentation automation.

Typically, the process starts with strategic discussions with the executive team, clinical and business leaders, and often a CFO or their team. Together, we identify the organization’s biggest operational bottlenecks and highest-value opportunities for improvement. One of our fastest-growing areas is ambient documentation, which we’ve integrated with Epic, Cerner, Meditech, and around 30 mid-market ambulatory EMRs. This allows us to streamline documentation workflows across specialties, acute care, ambulatory settings, and even operating rooms, freeing doctors from administrative burdens.

The typical rollout process starts with pilot deployments in select departments, working closely with clinicians to fine-tune models and workflows for their specific needs. Once the solution proves successful—often within six months, delivering time savings, cost reductions, and lower denial rates—we expand into other areas of the business.

Next, we often optimize revenue cycle operations or enhance patient-facing front-office functions, such as appointment intake and reminders. Our Commure Engage platform now facilitates nearly 150 million patient touchpoints per year, bridging the gap between front-office staff, clinical care teams, and back-office operations.

Finally, we help reimagine the billing stack and back-office workflows, addressing one of the biggest cost drivers in healthcare. By integrating across the entire patient and claim journey, we create a seamless, more efficient system that reduces overhead and improves financial sustainability for health systems.

We have a deep respect for Palantir and the company they

ve built. It took decades to reach where they are today, and their approach has demonstrated a few key lessons that we’ve embraced.

First, software markets are often much larger than people assume. Second, the traditional belief that software should be built once and never adapted for different customers is outdated. In reality, a lack of flexibility becomes a major handicap for businesses that don’t adapt to their customers’ unique needs.

At Commure, we embrace customization. No two health systems operate exactly the same way, and every deployment has unique requirements. But rather than seeing that as a challenge, we view it as an opportunity—every implementation teaches us something new, making our overall product experience better for all customers.

This approach also shapes the kind of engineers we hire. Our forward-deployed teams aren’t just writing code and handing it off—they’re on-site, engaging directly with customers, experiencing both the pain points and successes firsthand. They work closely with clinical leaders, helping prioritize problems in real-time. Health systems have 

 of issues to tackle, and if you try to solve all of them at once, you’ll fail. Success requires strategic prioritization—balancing guidance from HQ with the judgment of engineers and team leads on the ground who are earning customer trust.

We also aim to blend the best of both worlds. We have a strong foundation of ready-to-go products and infrastructure—solutions that already support millions of patients and tens of millions of appointments annually. But we also have the flexibility to customize and forward-deploy when needed. This combination is powerful: we’re not building from scratch with every new customer, but we’re also not forcing them into a rigid, one-size-fits-all solution.

From a business perspective, this leads to long-term software scalability and strong margins because of the repeatability of our core infrastructure. But in the short term, it ensures we’re solving the customer’s most pressing problems—because we have engineers on-site, actively engaged in delivering real solutions.

We’ve definitely benefited from the broader adoption of the Palantir model, which has made the concept of high-touch, custom-deployed software more mainstream—though it’s still relatively new in healthcare. The idea of engineers working closely with customers to build tailored solutions is gaining traction, and we’re seeing growing enthusiasm for that approach.

Where we provide the most value is helping health systems identify their top three automation or AI-driven opportunities—problems that large language models (LLMs) and automation can solve 

. Often, these are challenges they’ve discussed internally but haven’t formally analyzed with data-driven metrics.

For example, in revenue cycle deployments, we deploy a Financial Health Analysis (FHA) model with our RCM team that integrates customer data and helps pinpoint business gaps, such as, high denial rates in certain claim segments, claims bottlenecks that cause delays and leave money on the table, and poor documentation quality in specific subspecialties, leading to higher denials and extra workload for billing and coding teams

By starting with this structured analysis, we align our solutions directly with business pain points. This ensures that customers aren’t just deploying Ambient AI or automation tools because they read about them in an industry article—they’re using them because they directly address operational inefficiencies.

This targeted deployment strategy is critical. Many health systems have signed large contracts—sometimes $10 million or more—only to see no ROI because adoption was low. This happens when solutions aren’t properly integrated into hospitalist or emergency department workflows, where out-of-the-box Ambient solutions don’t work as seamlessly as they do in ambulatory settings. If you don’t forward-deploy and custom-integrate, physicians simply won’t use the product.

By prioritizing deep customer engagement, strategic implementation, and workflow-specific customization, we ensure that every deployment delivers measurable ROI within three to five months—not just in theory, but in practice.

It comes down to a few key factors. First, with any machine learning model, there’s an optimization function—you’re trying to optimize for a specific outcome. In an ambulatory setting, some physicians save a significant amount of time with Ambient because their notes come out perfectly. Others, depending on their documentation style or subspecialty, may not experience the same benefits. For them, quickly jotting down notes and entering structured data directly into the EHR is often more efficient.

Customization happens at the site level through deep engagement with the CFO’s department. We focus on areas where procedure capture rates are low, denial rates are high, and clinical disputes with payers are common. This is where Ambient can make a real impact, especially as we venture beyond core notetaking with extending assistive and agentic capabilities. By embedding care cues and nudges into the workflow, we help ensure that physicians document thoroughly and ask the right questions during appointments. This allows even the least-compliant or least-efficient clinician to document at the level of the best.

Within just a few weeks, we can see tangible improvements— documentation aligns more closely with RCM team requirements, coding queries from the back office decrease, and denial rates drop as a result. That’s our North Star for deployments.

In the emergency department, ROI is an especially complex challenge, and we’ve co-developed solutions on the ground with HCA to address it. For hospitalists, the challenge isn’t just capturing a single patient interaction—it’s synthesizing data from previous progress notes, discharge notes, chart history, family history, and medication history into a coherent action plan. Most Ambient vendors today have limited access to EHR data, relying mostly on audio input and a few high-level demographic fields.

With CommureOS, the depth of information we can pull is orders of magnitude greater. This enables us to generate significantly higher-quality notes, optimizing not only for physician time savings but also for RCM outcomes. The result is a solution that drives measurable ROI, both in efficiency and revenue cycle performance.

Yes, exactly. To your point, the sub-components of Ambient—such as care cues and the autonomous coding module—are modular functions that can also be integrated into other applications. From an engineering and design perspective, this modularity has been critical to how we scale our engineering team and expand our capabilities.

The biggest challenge in our business is managing the complexity of a large product suite. Internally, this means our engineers need to understand more than just the specific product they’re working on—they need broad context across the entire codebase and product suite. This is especially critical for forward-deployed engineers, who must be deeply familiar with the full stack. It’s not enough to specialize in Ambient just because that was their initial assignment at HQ.

To address this, we’ve built an operating model where engineers rotate through different products, ensuring they gain exposure to the full suite within their first six months. This cross-training is crucial, not only for on-site customer interactions but also for the ability to integrate and develop across multiple products seamlessly.

Our approach prioritizes modularity. Many customers start with just one product—like Strongline—but later want to expand into Ambient or revenue cycle solutions. The key is making it easy to deploy an initial product while ensuring that, when the time comes to expand, engineers on site can seamlessly activate additional apps using the same database access and EMR integration. That level of flexibility took years to build. The foundation of CommureOS was designed to balance modularity with interconnectivity, so customers can scale without needing a full redeployment each time.

We love taking on risk alongside our customers—not in the traditional insurance sense, but in that we’re willing to invest engineering and implementation resources upfront to deploy a solution for free, knowing that the ROI will materialize. We only start charging in a predetermined way once that ROI becomes clear.

Another key part of our strategy, similar to Microsoft, is aggressive bundling. If a customer is using Strongline, with physicians and nurses already wearing our badges and engaging with our app daily, launching Ambient is as simple as flipping a switch. That allows us to reduce the overall cost of transformation for a health system by offering a bundled solution from one vendor, rather than forcing them to piece together multiple point solutions.

In some cases, the effective cost of deploying Ambient is close to zero for a health system because they’re already using Commure Engage and Strongline. Since the incremental cost to activate Ambient is minimal for us, we pass those savings along to our customers.

Revenue cycle management is another area where our business model aligns closely with customer outcomes. In our midmarket business, we often structure our pricing around a percentage of collections in a cost-to-collect model or take a meaningful share of the improvement in collections or reduction in denial rates. This model ensures we’re fully incentivized to drive results, rather than simply charging a flat, million-dollar-a-year rate for software deployment.

Our relationship with General Catalyst is unique. Commure was incubated within GC, but beyond that, the connection mirrors the Peter Thiel-Palantir dynamic—Hemant is deeply invested in making Commure successful. At the same time, Commure serves as a nexus for many other GC innovations and companies, especially with what’s happening at Summa.

Our relationship with HATCo, the entity acquiring Summa, is more than just a partnership—we operate so closely that it often feels like one company. This level of integration allows us to make fast decisions, work directly with the teams transforming the health system, and collaborate with industry veterans to ensure Commure modules are truly impactful.

The setup is also structured as an at-risk model, meaning we align our success with measurable ROI. Given the scale and potential impact, Summa will likely become one of our biggest customers, on par with HCA. But because of our model, we can execute the deployment at minimal cost—something essential when working in a low-margin healthcare environment.

—like a Search and Summarization Agent that scours patient charts for relevant history, a Previsit Summary Agent that consolidates relevant lab data, and an MDM+ Agent that surfaces potential diagnoses and treatment pathways in real time. These solutions are live now.

We’re also introducing Voice Biomarker capabilities, integrating with devices like Apple Watch and “smart rooms” for continuous monitoring.

, which will further reduce manual coding tasks and accelerate reimbursement. All of these feed into our broader vision of “ambient everywhere,” where the technology quietly works in the background but has a transformative impact on clinical workflows.

At this stage, we can partner with health systems of all sizes, from the largest to some of the smallest. What makes a partnership truly effective, though, is alignment at the highest level—when leadership sees technology as a P0 priority for transforming their margin structure. If that imperative comes from the top down, things get done. Otherwise, projects tend to drag on indefinitely or become pilot engagements that never materialize into real impact. Every healthcare company has been through those, and they’re a waste of time for both sides.

The second key factor is alignment on the ground. The teams implementing our solutions need to be open to new ways of working—especially when we’re deploying multiple apps and transformations in parallel. This isn’t a one-and-done, three-month Epic IT project. It’s a continuous process with our team embedded on-site throughout the year, ensuring our modules reach the physicians effectively.

The final and most important factor is speed. One of Commure’s core values is speed—but not at the cost of rigor. There’s a common critique of tech companies that they 

 but that’s a misconception. We use the analogy of the Empire State Building, which was built in 400 days, versus the Millennium Tower in San Francisco, which took over five years and is now structurally unsound. The time it takes to build something doesn’t necessarily correlate with its quality. We hold our partners to that same mindset—moving fast and maintaining high quality are not mutually exclusive.

We have an excellent relationship with OpenAI, and based on every preview and product we’ve worked with, it’s clear they’re ahead of the pack. Recently, we’ve done frontier work with Google’s MedLM teams as well, and that has led to some key gains, in certain Ambient domains, for us. That said, we also very recently, we’ve shifted some innovation work to Anthropic models and developed our own homegrown models where it makes sense. We’ll always choose the best model for the given use case rather than committing exclusively to one AI stack.

The way we think about AI and language models is similar to electricity. I didn’t come up with this analogy—it’s widely used now—but what I like about it is that electricity didn’t create a trillion-dollar electric company. Instead, it created trillions in GDP growth and value that accrued to society. Of course, massive businesses were built on top of it—General Electric, for example, was a half-trillion-dollar empire in the 20th century. GE operated at both the application layer, making refrigerators and appliances, and at the infrastructure layer, powering the grid.

AI will follow a similar trajectory, with valuable companies emerging at every layer of the stack. We operate primarily at the application layer, using AI to build the best experiences and modules for our customers. But sometimes, we have to build parts of the “grid” ourselves— whether that’s fine-tuning models, developing custom training methodologies, or creating bespoke AI solutions. Wherever possible, we leverage and build on top of the foundational work of others, ensuring we stay focused on delivering tangible value to healthcare systems.

In many ways, the biggest risk and opportunity are two sides of the same coin—there’s an overwhelming amount to do in healthcare. When you look at our pipeline and the opportunities ahead, they feel almost infinite. We already have years of growth ahead of us with our existing partners. The key challenge is execution. With this much opportunity, anything short of building a multi-hundred-billion-dollar business would be an unforced error.

That pressure drives us—we’re perfectionists, and we’re building something designed to scale meaningfully within the U.S. healthcare system. What keeps me up at night is making sure we’re attracting and retaining the best talent across engineering, forward deployment, sales enablement, operations, and implementation.

Commure Scribe is an AI-powered medical scribe designed to automate clinical documentation for healthcare providers. It transcribes patient encounters, including multi-speaker conversations lasting over three hours, and converts them into structured notes such as SOAP and progress notes. The platform supports over 60 languages and offers customizable templates to fit various specialties and workflows. Additionally, it provides automated CPT and ICD-10 coding recommendations to enhance billing accuracy.

HITRUST CSF

HIPAA

SOC 2 Type 2

201-500

Ambulatory Practice

Hospital / Health System

Commure Scribe

Hospitals / Health System

Provider / Payer

The community health center selected the AI ambient scribe solution to roll-out across its clinic locations through an Epic integration.

Commure Scribe

Customers