Earnings season rolls on with Palo Alto Networks (PANW:NASDAQ) reporting this week. Cybersecurity has become shorthand for national resilience, and Palo Alto’s results will show whether governments and corporations are keeping pace with escalating threats. Last week’s technology news was dominated by Cisco’s strong quarter and Oracle’s push to reinvent the electronic health record. This week shifts focus back to the firewall.

As always, you can browse past notes and the searchable research archive at the Epistrophy website.

As always, I’m focused on three things:
1) Technology-driven change;
2) the latest in innovation and startup trends, and;
3) stock fraud.

Companies Discussed

In This Note:

• Companies Discussed

• Oracle’s AI Doctor Is In The House

• Tweet O’ The Week

• Epistrophy In The News

• 📆 of Epistrophy Events

• Availability This Week

Oracle’s AI Doctor Is In The House

Ambient AI EHR: Less Clicks, More Care

In American medicine, the electronic health record has become the symbol of a broken system. Doctors spend more time clicking boxes than talking to patients. Hospitals have digitized charts, but physicians remain secretaries. The EHR, intended to bring order, has instead become shorthand for burnout.

A July 2022 report, “Electronic Health Record Use and Family Physician Burnout in the United States” (JAMA Network Open), a cross-sectional study of more than 10,000 family physicians found that 42.8 percent reported symptoms of burnout, with after-hours EHR work nearly doubling the odds of exhaustion. In In April 2025 an influential finding published by the American Medical Association revealed that for every eight hours of scheduled patient care, physicians spend more than five hours on EHR documentation. “Prevalence of Burnout Associated With Electronic Health Record Use Among Health Care Professionals: Systematic Review and Meta-analysis”, a January 2024 pooled analysis across 66,000 clinicians reported an EHR-related burnout prevalence of 40.4 percent and an odds ratio of 2.43 for those spending extra hours in the system.

The policy backdrop goes back to Washington. The HITECH Act of 2009, championed by then–Vice President Joe Biden and passed as part of the Obama administration’s stimulus, offered $27 billion in incentives for “meaningful use” of certified EHR technology. The law worked in the narrow sense: hospitals digitized their records. Federal officials promised as much as $27 billion in incentive payments and the Congressional Budget Office estimating total Medicare and Medicaid outlays closer to $32.7 billion over 2009–2019 (CMS; Commonwealth Fund; CRS/CBO). The money worked as a lever. Adoption among U.S. hospitals climbed above 90% within a decade (HealthIT.gov). 

But the design lagged behind the mandate. Clinicians found themselves doing clerical work in systems built for billing and compliance. Inbox messages multiplied. Alert fatigue set in. Burnout became measurable, then chronic.But it entrenched legacy vendors, locked in billing-first architectures, and left physicians to navigate clunky interfaces never designed for clinical care.

At the time, Cerner needed a stronger hand. It was losing ground to Epic, which has become the default for large health systems. In 2024 Epic’s hospital share reached 42.3% while Oracle Health’s share fell to 22.9%, with a net loss of 74 hospitals and 17,232 beds that year, according to KLAS. KLAS tied some defections to “declining performance and ongoing revenue cycle issues” on the Millennium platform. The optics weren’t helped by high-profile problems in the Department of Veterans Affairs rollout, where watchdogs tallied more than 800 major incidents tied to the new EHR since go-live. Those are solvable engineering problems, but they hardened a market perception: Epic wins the largest enterprises, Cerner defends the rest 

It is into this landscape that Oracle has chosen to march. Last week, the company unveiled its rebuilt electronic health record platform, the most visible product yet from its $28.3 billion acquisition of Cerner, a deal announced in December 2021 and completed in June 2022. Oracle paid cash, not stock, positioning the purchase as a way to bring its database and cloud muscle into one of the few industries where software adoption remains uneven. At the time, Cerner was struggling with slowing growth, declining hospital contracts, and a reputation for poor usability. Epic dominated new installations; Cerner was losing ground. Oracle framed the acquisition as a turnaround play— founder and Chief Technology Officer Larry Ellison called health care “a mission” and promised to build “a unified national health records database.”

That ambition now has its first product. The new Oracle EHR is pitched as conversational, cloud-native, and embedded with AI. Ellison promised that physicians will be able to speak naturally, with the system generating structured notes, orders, and billing codes automatically. In his words, the EHR will become “an assistant, not an adversary.” 

Seema Verma, who runs Oracle Health and Life Sciences, told us that the company resisted bolting new features onto old code and built an “entirely new EHR” for the agentic AI era. The company calls out dozens of native AI agents that share context and orchestrate workflows across care, operations and reimbursement. It emphasizes explainability and the ability for clinicians to challenge an agent’s reasoning. The system is “not a walled garden” — customers can extend Oracle agents, build their own or integrate third-party models while keeping workflows patient-centric. This is straight from the launch materials and the press release, which also promises a consumer-grade experience that “brings back the joy of practicing medicine”

Oracle says its semantic data model will allow records to interoperate across institutions, ending the fragmentation that has plagued American health IT since its inception.

The technical claims matter. Today’s EHRs remain walled gardens. Epic’s system dominates large hospitals but is famously closed. Cerner, before Oracle, lagged in interoperability. Doctors switching hospitals often find that patient histories cannot follow. Oracle is promising a universal patient record, structured on open standards, queryable in real time. It says it will deploy generative AI models to surface clinical suggestions, flag gaps in care, and summarize histories at the point of care.

The market backdrop is large. Health IT spending in the U.S. exceeds $40 billion annually. Epic remains privately held and entrenched. Oracle, as a public company, is wagering that the combination of its database stack, its Gen2 Cloud, and its newly rebuilt EHR can win share in a sector that has resisted outside disruption. For hospitals, the promise is lower IT overhead by running records in Oracle’s cloud, integrated analytics, and less time spent hiring scribes to chase physicians’ documentation.

Ellison has framed this in moral terms. “Because the pandemic has shown a variety of weaknesses in our healthcare systems,” he said on an earnings call in March 2022, “and now we have the technology to address those weaknesses, and that’s what we’re going to do.” It was a characteristically blunt expression of a database man’s faith: that information, structured and available, can fix what regulation and habit have not.

The problems Oracle faces are not only technical but cultural. Physicians do not trust corporate promises. The memory of Meaningful Use lingers. Doctors were told then that EHRs would make care safer and more efficient. Instead, the systems became billing engines. Burnout rose. Medical errors persisted. Surveys show that more than 60 percent of physicians say current EHRs detract from their clinical effectiveness. Convincing them that a new Oracle-built Cerner will reverse that is a task of credibility as much as code.

Epic remains the elephant. Its market share is north of 35 percent among U.S. hospitals. Its customers are loyal. Oracle must persuade large hospital systems to rip and replace, or at least to let Oracle’s cloud sit on top. The company says it will pursue both: migrate Cerner’s base to its new EHR, and offer interoperability layers to Epic clients who want AI-driven summaries and analytics without switching.

Financially, the Cerner deal was the largest in Oracle’s history. At $28.3 billion in cash, it was a bet the company could extend its enterprise software dominance into health care. At the time, Oracle’s core database business was slowing. Cloud infrastructure was its growth story, but it lagged Amazon (AMZN: NASDAQ), Microsoft (MSFT: NASDAQ), and Google’s parent Alphabet (GOOGL: NASDAQ). Cerner offered not only revenue—$5.8 billion in 2021—but an industry vertical that had resisted those rivals. Oracle said then that Cerner would be “a huge growth engine for years to come.”

Three years later, the skepticism remains. Oracle’s applications revenue has grown, but Cerner has been a source of R&D, not revenue growth. And it’s been a tough business. Hospitals, pressed by inflation and post-pandemic labor shortages, have been cautious to sign new IT contracts. Epic has continued to win the marquee deals, including a string of academic medical centers in 2023 and 2024. Oracle’s launch last week is, in part, a bid to reset the narrative: to show that Cerner was not a stranded asset but the foundation for a new product cycle.

The technical details deserve scrutiny. Oracle says its EHR is built on a semantic model that uses standard ontologies like SNOMED CT and LOINC, allowing structured data exchange. It claims natural language processing models will parse physician speech in real time, generating both structured fields and narrative notes. It says orders will be executable by voice. The company has demonstrated mock-ups of a physician saying, “Schedule an MRI of the lumbar spine and start the patient on 20 milligrams of atorvastatin daily,” with the system producing a compliant order set, checking for drug interactions, and scheduling imaging without further clicks.

Skeptics will note that these capabilities have been promised before. Nuance, now part of Microsoft has long offered clinical dictation. Smaller startups have built ambient AI scribes. None have scaled across major hospital systems, in part because error rates are unacceptable in clinical care. Oracle insists its models are more accurate, trained on Cerner’s massive dataset, and integrated directly into the workflow.

Interoperability is another test. The U.S. has spent billions on health information exchanges, with limited success. Oracle says its EHR will allow physicians at different hospitals to query the same patient record. It promises that if a patient seen at Mayo Clinic turns up at Stanford, the attending physician will see a unified history. Achieving that requires not only technical standards but contractual agreements between health systems that have historically guarded their data.

For patients, the promise is appealing. A single longitudinal record could reduce errors, eliminate duplicate tests, and empower individuals to access their own histories. For physicians, the test is more pragmatic: does it reduce clicks? Does it save time? If Oracle can cut documentation hours by even 20 percent, it will have a case.

The economics matter because hospitals are under strain. Operating margins remain thin post-pandemic. Labor costs are high. An EHR that claims to improve physician productivity and reduce IT overhead must deliver measurable savings. Oracle says its cloud infrastructure can lower costs compared to on-premise data centers. It pitches its AI as reducing the need for human scribes, a market that has grown precisely because doctors cannot keep up with documentation.

The politics matter too. Health IT is regulated. The Office of the National Coordinator for Health Information Technology sets certification criteria. Congress continues to push for interoperability. Any national record will face privacy questions. Oracle has promised that patient data will be secure, encrypted, and accessible only with consent. But its reputation is not built on patient trust; it is built on enterprise contracts. Epic has long marketed itself as physician-friendly, even as doctors complain. Oracle must overcome the perception that it is an outsider.

What does success look like? If in five years Oracle has stabilized Cerner’s base, migrated a majority to its cloud-native EHR, and signed even a handful of Epic clients to its interoperability layer, it will have changed the competitive balance. If not, the $28.3 billion will look like an expensive misstep.

The broader significance is that a major enterprise software company is betting on fixing one of medicine’s most persistent failures. The EHR was supposed to be the backbone of modern care. Instead, it became a billing platform that doctors resent. Oracle is promising to rewrite that story. The launch last week is the beginning of the test.

If it works, physicians will spend less time clicking and more time talking to patients. If it fails, burnout will remain the operating system of American medicine, and Oracle will have joined the long list of companies that underestimated the culture of health care.

Tweet O’ The Week

Epistrophy In The News

On Schwab Network, I broke down Cisco’s latest earnings, highlighting the company’s robust demand from hyperscalers and early signs of revenue from its Splunk acquisition (I’m open to commentary about that shirt). The discussion centered on whether Wall Street is underestimating Cisco’s role in the enterprise AI build-out and whether its security business can compete with Palo Alto Networks. We may see this week!

📆 of Epistrophy Events

Availability This Week

I’ll be in Silicon Valley and Los Angeles meeting with clients this week. These trips are a reminder that the technology conversation happens as much in boardrooms as in earnings calls. If you’re nearby, let me know.

Written reports are available to clients, with video summaries on YouTube, and of course our popular summaries of the summaries on Instagram, TikTok, and YouTube Shorts.

Are these notes are helpful to you? Suggestions? I’d love to discuss them further and, as always, comments, questions and ideas are appreciated.