Team Viz
How Generative AI is Redesigning Clinical Workflows
Q: Andrew, it’s great to interview you. I feel like we have a kinship – we both joined Viz.ai this year, come from clinical backgrounds, and are energized by this next chapter of the company’s growth. To start, I’d love to hear what drew you to join Viz?
A: My career has been focused on the best ways to deliver care for patients. I live that in the OR and clinic every week, but as we both know, there are so many levers that help or hinder our ability to provide that care. I’ve been a health services researcher and poured my soul into 200+ papers looking for all the different things we can measure and influence, and then spent 6 years as a chief medical officer of a global architecture literally trying to design the physical hospital to deliver better care. While I think those have led to important improvements, I see the rise of generative AI as a once in a generation opportunity that outscales anything else I’ve done. Once I saw the early demo and met the team behind it, I knew it was something special I wanted to be a part of. The reality is – everything in the hospital is harder than it needs to be. And if we don’t make the system simpler, easier to navigate, many of our colleagues will leave clinical medicine and more patients will struggle to find doctors.
Q: That’s inspiring. You recently co-authored an article in NEJM AI with Chris Mansi that painted a stark picture of the pressures facing health systems. With that in mind, what were the big challenges you wanted to tackle with Viz Assist?
A: The problem is simple but devastating: clinicians spend nearly twice as much time in the electronic health record (EHR) as they do with patients. That means hours of late-night “pajama time,” burnout, and gaps in care because important details get lost in hundreds of pages of records. I know these doctors personally, and many don’t eat dinner with their families because of the time they spend chart prepping. Viz Assist was created to give that time back.
Q: You’ve mentioned Viz.ai’s evolution a few times — why was now the right moment for the company to go in this direction?
A: We’ve always been the system of action. Today, more than 1,800 hospitals across the country rely on Viz.ai’s AI-powered detection, real-time imaging, measurements and care team coordination to accelerate time to treatment. For non-urgent disease we automate referral workflows to reduce the number of patients lost to follow-up. We pioneered the use of AI embedded into clinicians’ workflow in healthcare long before it became a buzzword. So it only makes sense that we’re continuing to provide tools that help physicians act and remove the friction from care. Whether it’s instantly coming up to speed on a patient, streamlining documentation, or automating coding.
Q: There’s been a flood of generative AI tools hitting healthcare lately. What makes Viz Assist stand out from the rest?
A: A lot of AI tools, including those within EHRs, are generic note generators or aggregators. Viz Assist is different — it creates condition-specific, specialty-aware summaries that link directly to the source data. What a cardiologist looks for is very different from what a neurologist screens for and Viz Assist knows that. It also integrates imaging, notes and EHR data together, so we can surface things that otherwise get buried — like a missed anticoagulation or a patient who should have been referred for defibrillator placement. It doesn’t just produce text; it distills what matters for that clinician, in that moment. It feels like having an expert chief resident with you all the time.
Q: A lot of organizations already use ambient documentation tools. In that case, why would they still consider Viz Assist?
A: Viz Assist doesn’t replace your ambient tools; it integrates with them. If you’re happy with your vendor, we plug them into Viz Assist’s workflow. If not, we can provide our own or recommend alternatives. The value comes from unifying data sources into a single, seamless workflow for clinicians. So the imaging inputs, EHR summarization, ambient recordings and billing recommendations all integrate into one place, Viz Assist.
Q: Integrations always raise eyebrows with IT teams. How hard is it really to bring Viz Assist into an existing EHR environment?
A: Many health systems worry about integration being burdensome. Viz has EHR integrations at over 100 sites and built a streamlined process to minimize IT lift. While EHR integration is required for Viz Assist, our dedicated team ensures it’s straightforward and low-burden for your IT department. We just had a system complete their EHR integration with us in under a week. We move as fast as the hospital is able to, and our dedicated implementation is ready to support every step of the way. One thing we love to do is an early IT engagement, well before a contract. Often with just that meeting we can give a realistic expectation of what is involved and how long it will take. We almost always find that after we meet with IT and they see what is needed, they tell us it’s an easier lift than they anticipated.
Q: Healthcare data comes in every imaginable format. Can Viz Assist actually make sense of all that — from PDFs to imaging to structured data?
A: Yes. If clinicians need to dig through it today—whether structured EHR fields, imaging, or scanned PDFs—Viz Assist can extract and surface the relevant details. That flexibility is what makes it so impactful across specialties and service lines. One of our favorite aspects of it is not just the extent of content, but the indexing. So if Viz Assist tells me there was an operation 10 years ago that it found in a scanned PDF, it can hyperlink me to the primary content. So in a way, it’s both an essential summary but also a table of contents for places where we may need a deeper dive.
Q: Let’s bring this to life. What does Viz Assist actually look like in action for a clinician during their day?
A: You’re an oncologist, Tim, so let’s use you as an example. Imagine walking into clinic and instead of spending two hours piecing together records, Viz Assist has already done the work – summarizing tumor markers, prior treatment, and imaging in a concise timeline. During the visit, it aims to listen, draft the note, and over time expand to suggest billing codes tied to evidence in the chart. You’re still the editor, but you’re no longer starting from scratch. That’s what makes it different: it’s not just insight, it’s actionable.
Q: Beyond the clinical side, why should health system leaders care about this? What’s the bigger picture impact?
A: Documentation isn’t just a clinical issue — it’s financial and operational. Missed codes or incomplete documentation can lead to denied claims, lost referrals, and unrealized revenue for care delivered. With Viz Assist, clinicians can practice at the top of their license. Too many doctors, PAs and nurse practitioners are spending hours pre-charting when instead, they could be with patients. Viz Assist is a rare double win: it gets clinicians back to the work they want to do while strengthening the business and clinical excellence of the health system.
Q: I know you’ve started seeing real-world results already. Can you share some examples of what that impact looks like so far?
A: Early adopters are seeing big gains. At one site, cardiology chart review time was cut from 45 minutes to 5. At another, a neuro clinic using Viz Assist identified 4 patients that would’ve otherwise gone missed within 1 week of using the tool. Those are real patients, real impact. And this is just the beginning.
Q: And looking ahead — where does Viz Assist go from here?
A: Right now, we’re supporting specialties like cardiology, neurology, oncology, and pulmonology. But the vision is much broader — an intelligence layer that underpins the entire health system. Imagine a tumor board prep automatically generated, or a population health manager instantly identifying thousands at risk. That’s where this is headed.
Q: Before we close, security and compliance are always top of mind. How does Viz Assist handle patient privacy and HIPAA safeguards?
A: Security is foundational. All data is encrypted, access-controlled, and fully HIPAA-compliant. Clinicians can collaborate securely without resorting to unsecure texting or emailing, ensuring both safety and efficiency.
Q: Finally, what’s the one thing you want clinicians — and health system leaders — to take away about how generative AI can impact their work?
A: It’s the help you have wanted for a long time. We all know too many of our colleagues who want to quit clinical medicine because they’re overwhelmed by the day-to-day administrative burdens – pre-charting, documenting, care-coordinating, billing – that pull them away from true patient care. Generative AI is redesigning workflows so that the right information surfaces at the right time — freeing clinicians to focus on what only humans can do: caring for patients.
Learn more here.