Sponsored

Art and Science: Building Intelligence at the Point of Conversation

Advertisement

Health system innovation moves at the speed of trust

In healthcare, the real power of artificial intelligence won’t come from data alone, but from dialogue.

At the 2025 Becker’s Health IT Conference in Chicago, Abridge brought together a cross-section of healthcare’s clinical and technology leaders for Art and Science: An Executive Summit on Intelligence at the Point of Conversation. In a year when AI’s promise feels both exciting and uncertain, the summit explored how real intelligence is built — not in the lab, but where human connection meets data.

The gathering opened with Abridge CEO and Co-Founder Shiv Rao, MD, who framed the current moment in AI’s acceleration with candor. “At this intersection of healthcare and technology, it feels like the ground is shifting so quickly,” he noted. “We’ve got this incredible opportunity to build the right things in the right order and then to measure them responsibly and then to scale them.”

The conversations that followed moved far beyond product talk. They traced how ambient intelligence — AI built to listen and reason within the clinical conversation — has begun to transform healthcare from the inside out. Across payers and providers, front-line leaders described a shift from automation to empathy, from documentation to decision-making, and from isolated innovation to shared trust.

What follows are six central takeaways from Art and Science: An Executive Summit on Intelligence at the Point of Conversation.

1. Healthcare is being redefined by intelligence that begins at the point of conversation.

Every medical episode begins with a story: a patient describing what they feel and a clinician interpreting what it means. Dr. Rao posits that this story remains the most valuable and underutilized data source in healthcare.

Founded in 2018, Abridge has partnered with more than 200 health systems across the U.S. Its ambient AI and large language models capture those stories, distill them, learn from them, and translate them into structured intelligence that powers more connected care. Dr. Rao described Abridge’s roadmap across three dimensions: before, during, and after the patient-clinician conversation; depth across specialties and settings; and innovation that bridges business models. Each dimension reflects Abridge’s conviction that healthcare intelligence begins at the source — in the dialogue itself.

“We see our jobs as equipping clinicians with superpowers to help them be the best possible advocates, caregivers and healers,” Dr. Rao said.

Abridge’s contextual reasoning engine synthesizes conversational data with information from the chart, specialty-specific models, and payer rules to generate documentation that is simultaneously clinically useful, billable, and compliant. Dr. Rao previewed that the company is investing a great deal of time and energy on advancements in the inpatient setting and clinical decision support, marrying the science of LLMs with the art of clinical reasoning.

2. Payers and providers are using AI to replace friction with transparency and trust.

Richard Clarke, Chief Analytics Officer at Highmark Health, joined Dr. Rao to discuss a collaboration aimed to transform one of healthcare’s most frustrating experiences: prior authorization.

The payer-provider interface has been marred by friction for decades: duplicative effort, black-box rules, patient frustration. Mr. Clarke framed the stakes clearly. “There’s a lot of pressure, and that performance pressure can lead to a number of different things,” he said. “I’m hoping it leads to increased collaboration that keeps the member and patient at the center, and leads to better health outcomes and more affordable care — not the hellscape of warring AI on each side that just weaponizes the payer-provider interface.”

Highmark and Abridge are developing an AI-driven, real-time prior-authorization workflow that brings transparency into the exam room itself. As the clinician talks with the patient, the system automatically identifies the relevant CPT code, determines whether prior authorization is required for that patient’s plan, retrieves the applicable criteria, and instantly displays which requirements have been met.

Mr. Clarke elaborated: digitized guidelines could appear as five gray Xs turning to green checks as each requirement is met. “It’s kind of a beautiful, simple vision. But what’s required behind that — not only the AI in the conversation, but a payer essentially exposing a lot of their incredibly legacy architecture out to the models — is really what we need to work on to make it real,” he said.

That transparency, both Mr. Clarke and Dr. Rao agreed, is ultimately about trust between payer and provider, clinician and patient, technology and user. As Dr. Rao reflected, the challenge isn’t efficiency — it’s rebuilding trust, which is in shorter supply in healthcare and beyond.

“We’re trying to rebuild that trust,” said Dr. Rao, among clinicians and patients but also within healthcare more broadly. “It’s about being transparent, but also giving that clinician, at the point of the conversation, the psychological safety that we’re not going to change the guidelines on them. And if these boxes are checked off, you truly will get that approval and be able to tell the patient they can move forward.”

3. The new ROI of ambient intelligence: time, money and lives.

Highmark’s prior authorization work offers one glimpse of how generative technology could reshape healthcare’s economics. “I’ve never seen anything that’s more tailormade for an industry,” said Mr. Clarke. “It just seems to be pointed at so many of the challenges that we have, across both payer and provider.”

For Dr. Rao, that alignment is no accident. From its earliest days at UPMC, Abridge was designed not only to generate drafts of clinical notes but to understand the conversation itself, the narrative core of care that clinicians rely on and health systems bill from. Years of fine-tuning have given Abridge a rare vantage point inside the flow of care itself: close enough to where the story begins and early enough to influence where it leads.

The returns, Dr. Rao said, follow a deliberate arc. Time is all about returning hours of focus and presence back to clinicians, including reduced pajama time and note completion time. From there comes operational efficiency—ensuring that every note is not only clinically rich but also billing-compliant, with more-accurate documentation supporting long-term sustainability.


And at the top of that arc are lives. The same conversational foundation that creates accurate notes can also fuel real-time decision support, smarter risk management, and more responsive models of care. “How do we actually improve outcomes — not just the patient experience — but how do we actually save lives?” Dr. Rao asked.

4. Health systems are proving that AI adoption succeeds when it starts with people, not products.

If Highmark’s conversation traced the science of intelligence, the next discussion turned to its art — the deeply human work of change management. Across UPMC, Duke Health, Reid Health and Endeavor Health, leaders described how ambient AI was reshaping their organizations not through software alone, but through enthusiastic adoption and renewed trust and joy in practice.

At UPMC, where Dr. Rao still rounds as a cardiologist, Chief Technology Officer Chris Carmody reflected on the system’s role as one of Abridge’s first partners. “We knew the technology was special,” he said. “Even though Abridge was born here, we still put them through the ringer.” UPMC’s 42 hospitals are now moving toward a single digital foundation, with plans to expand Abridge to 12,000 physicians this fall.

And it doesn’t stop there. The health system is actively seeking out other problems to solve with the ambient AI technology. “We have a whole plethora of items that we want to do in leveraging the technology we’re talking about with Shiv and the team at Abridge,” Mr. Carmody said. “We see this as a great partnership.”

At Richmond, Ind.-based Reid Health, Chief Nursing Officer Misti Foust-Cofield saw the momentum of Abridge leap from physicians to nurses nearly overnight. “We noticed quickly how wildly successful this was on the ambulatory side, and I thought, we need this for nursing,” she said. “So we quickly engaged with Abridge, with the focus truly being on human connection and bringing the joy back to the work at the bedside. Much of nursing is about the art of nursing and the science of nursing, but Abridge has allowed us to really focus on the joy in nursing.”

For Endeavor Health, born from the merger of four Illinois systems, the challenge was unity. Physicians were still adjusting to a shared identity, said Justin Brueck, system vice president of innovation and research. When identity is unsettled, innovation can be harder to embrace, much less scale.

“We had to get them excited about joining and basically giving up the autonomy of their own brands, their own communities, their own ways of working,” he said. “If we had not been able to solve for the burnout factor or their overall experience, my job would right now be even harder because there would not be a willingness to see innovation come to the table.”

Within months, the results were visible at Evanston, Ill.-based Endeavor: clinicians reported burnout symptoms dropped 40%, and physicians reporting adequate face time with patients rose from 50% to 80%. Adoption patterns also defied assumptions.

“People said that our older clinicians were going to be a little bit hesitant to use this, and actually over 30% of our users are 20 years-plus in practice,” Mr. Brueck said. “And that’s where you get those stories of, ‘I’m going to practice for another five years because of this.'”

5. Partnership is the new measure of innovation.

For all the talk of algorithms and platforms, the conversations kept returning to something more human: the role of partnership in making innovation real.

At Duke Health, Senior Consultant for AI and Technology Engagement David Claxton described collaboration that felt less transactional than communal — a rare dynamic in an era of AI overpromises and vendor fatigue. “Even when some of the specialists who think the notes could be tailored more toward their specialty, the partnership that we’ve seen with Abridge has been something that we haven’t experienced before,” he said. “It really feels like a true partnership with them versus a commercial contract.”

That sentiment echoed across systems. At Reid Health, Ms. Foust-Cofield said Abridge has been open-minded to hearing and fielding all ideas from her and her team, even those that can sometimes seem wild at the start. “They’re incredibly open to us pushing them and to them pushing us, which makes us better,” she said.

For Mr. Brueck, the feedback loop has become a lifeline. He recalled clinicians’ buy-in even during the training period. “We had webinars that physicians would stay on beyond the half hour, beyond the hour — to the hour and a half — just because they got more and more excited about it,” Mr. Brueck recalled. “The Abridge team was there to listen to them; they didn’t just hang up and say, ‘Next call.’ It wasn’t about selling, it was really about people who were excited about the solution and what it could do.”

As Dr. Rao noted, Abridge’s ethos begins and ends with that loop: Feedback is oxygen. “It’s literally what keeps us going every single day.”

6. Cultural change today is shaping the care models of tomorrow.

Adoption is never frictionless. At Duke and Endeavor alike, surgical specialists lagged behind primary care in uptake; some clinicians considered themselves “too efficient” to need help. Leaders soon realized that success depended less on convenience than on conviction. By focusing on quality and compliance — showing how AI-generated notes captured higher-value encounters and improved documentation integrity — they began to shift perception from novelty to necessity.

At Reid Health, Ms. Foust-Cofield went further, recasting implementation itself as a cultural movement. “Technology can be a four-letter word,” she said. “Our approach to Abridge has been more of a cultural initiative than a technology initiative. It’s something we want to bring to our patients and caregivers, not a tool they have to reshape their process around.”

As those cultural initiatives take hold, leaders took stock of what comes next: not just faster documentation but a reimagining of care. Mr. Brueck of Endeavor Health called for “first-principles thinking.” Rather than layering AI onto old workflows, he urged leaders to redesign care around new possibilities — using ambient data to segment populations, support prevention and accelerate clinical research.

“Most clinicians, when they came out of medical school, they had this little flame inside of them that said, ‘I want to change the world,'” Mr. Brueck said. “By us doing these things and creating the brain width, the mind width, all those things, they can actually start to lean into the future of healthcare again.”

Mr. Carmody of UPMC envisioned a simplified, device-free hospital: “Let’s get rid of all the WOWs (workstations on wheels) and PCs everywhere. With the right AI and mobility, clinicians can focus fully on patients, not screens.” Ms. Foust-Cofield imagined smart-room integration and procedural settings where real-time voice documentation could capture every action, ensure compliance, and improve billing accuracy.

Threading through each vision was Dr. Rao’s metaphor of air-conditioning. “One mantra that we’ve had is good air conditioning—where the user experience should feel like that when it’s set right, the temperature, you’re not aware of it. You’re just comfortable. Similarly, if we can be invisible here, that would be amazing.”

Conclusion

Art and Science: An Executive Summit on Intelligence at the Point of Conversation was less a celebration of technology than a reflection on its place in care. Across every conversation, a common thread ran through: intelligence in healthcare begins in the exchange between patient and clinician.

Trust, not speed, is proving to be the real foundation for progress, and time the first return on investment. Leaders in the room showed that meaningful change depends less on software than on people and cultures willing to evolve and partnerships built to last. Ambient intelligence, in that sense, feels less like a breakthrough than a recalibration — a way to restore focus, connection and purpose to the work of care itself.

Dr. Rao noted that, in many ways, these are still the early innings for what ambient AI can unlock, how far upstream its intelligence can reach and how deeply it can personalize and adapt to each clinician.

“We recognize that we need to go millions of miles deep — and part of millions of miles deep, we think, is personalization,” he said. “So a big push for us now that we’re at this scale is not just making sure that the floor meets a certain standard across all different specialties and all different clinicians out there, but that we can get up to the ceiling of whatever an individual might want.”

Advertisement

Next Up in Health IT

Advertisement