Priming the Ambient AI Medical Scribe with Skriber:

Insights from Dr. Connor Yost’s Interview with Healthcare IT Today

In a recent interview with Healthcare IT Today, Chief Medical Officer Dr. Connor Yost shared the full story behind Skriber, the ambient AI medical scribe that is quickly gaining attention across healthcare. The conversation revealed not only how the product works, but why it was built, the deeper problems it aims to solve, and why physicians are responding so strongly to it. Through his own experience as a clinician and AI researcher, Dr. Yost described a tool designed with a simple goal: reduce documentation pain without forcing doctors to change who they are.

A Clinician Who Understood Both Medicine and AI

Before joining Skriber, Dr. Yost was deeply involved in studying large language models and their impact on healthcare. He had earned AMA innovation grants and had published research on clinical AI. He was also an early adopter of AI scribe technology during his training. But those early tools left him frustrated.

Doctors were excited about AI scribes, but the excitement faded when the notes arrived. They looked messy, unstructured, and incomplete. He explained that many of these tools “didn’t feel like a physician was involved in creating them,” because on the surface they looked impressive — but when it came time to use them, physicians ended up fixing and rewriting large portions of the note.

He pointed out a major truth:

Only about 30% of a clinical conversation becomes part of a note. The other 70% is what the doctor is thinking, analyzing, and processing internally.

No AI system was capturing that 70%.

This gap between medical thinking and AI output became the foundation of his motivation.

The Co-Founders and Why Skriber Exists

The story of Skriber began when Dr. Yost met co-founders Hayden Plummer and Matthew Priest, two experienced entrepreneurs who had previously built and exited a renewable energy technology company. After their exit, they wanted to take on a meaningful problem. Both men had watched physicians in their families struggle with long hours and overwhelming documentation demands.

They kept hearing stories of doctors staying late at night to finish charts. They saw the emotional exhaustion that came from this workload. They wanted to build something that eased that burden.

What surprised Dr. Yost was how humble the co-founders were. He said many successful tech leaders act as if they already know everything, but Hayden and Matthew were different. “They listened,” he said. “They didn’t want to tell me what the product should be — they wanted me to tell them what it needs to look like.”

This humility, paired with strong engineering talent, created a collaboration that felt natural. Together, they formed the vision for Skriber:
a medical scribe built around real physician needs, not tech assumptions.

Why Early Versions Failed — And What Skriber Changed

During early development, Skriber’s engineers built a prototype based on other scribes in the market. They were excited to show it to Dr. Yost. His reaction was immediate: “This is horrible.”

The sections were wrong. The clinical flow was wrong. It didn’t sound like a real physician note. That moment forced a reset in thinking. The team realized they needed to design the product from the doctor’s perspective, not from the perspective of other software tools.

From that point forward:

• physicians shaped the templates
• physicians shaped the prompting logic
• physicians shaped the structure
• physicians shaped how ICD-10 codes were captured
• physicians shaped the style and tone

Skriber also changed onboarding. Every new doctor works directly with an engineer so the tool can match their personal documentation style. As Dr. Yost said, “We hear the physician’s voice.” That voice is then used to “mimic and replicate their style of writing,” because the doctor is the one who signs the note — not the company.

The Most Important Insight: Priming the Scribe

One of the strongest themes in the interview was Skriber’s honesty about behavior change. While many AI scribe companies claim “zero behavior change” is required, Skriber takes a more realistic and transparent approach. A very small behavior change is required — but the payoff is enormous.

Before walking into the exam room, the physician gives Skriber a short overview of the patient. Just 10–20 seconds. This includes:

• the patient’s medical history
• the diagnosis
• the reason for today’s visit
• any important background

According to Dr. Yost, this changes everything:

“The quality of the note is astronomically higher when physicians prime the scribe.”

Without priming, the AI is forced to fill in gaps and make assumptions. With priming, it understands the full context before the conversation starts. This single step improves accuracy, reduces hallucinations, and produces more complete notes while keeping the visit natural.

Helping Doctors Avoid Costly Billing Mistakes

Documentation is not just about clinical accuracy — it affects billing, insurance approval, and workflows across the care team. One of the most common issues physicians face is failing to use the exact diagnostic term required for reimbursement.

Dr. Yost gave a clear example:

A doctor may describe every sign of sepsis — hypoxia, elevated lactate, pneumonia — but if the note never explicitly states “sepsis,” the claim may be denied.

Skriber automatically identifies the medical story the doctor is telling and helps map it to the correct ICD-10 codes. It ensures the clinical language supports the billing requirements. As Dr. Yost explained, the problem is often not that the physician did something wrong; they just “picked the wrong problem” or didn’t use the exact term needed.

Skriber bridges that gap.

A Workflow Designed Around Real Practice

During the interview, Dr. Yost described what the typical workflow looks like. Physicians can use Skriber through a mobile app, web browser, or computer. They start by priming the note. Then they record the visit as usual. They can also add comments after the visit for sensitive details they did not want to say in front of the patient.

The physical exam is handled through structured templates. Physicians only mention what is abnormal. Skriber fills in the normal findings based on specialty. A doctor might simply say, “You sound a little wheezy on the right side,” and Skriber automatically converts that into proper clinical terminology.

The note is ready in 2–3 minutes, and doctors can paste it into the EHR or let it auto-insert if integrated.

Accuracy and Honesty About AI Limitations

Dr. Yost offered realistic numbers about transcription accuracy. Dragon dictation sits at about 3–5% error. Skriber is currently at 6–7%, with the biggest variable being microphone quality. He expects this number to drop as device hardware improves.

More importantly, Skriber includes guardrails to prevent the AI from guessing or inventing details. If the system is unsure, it asks for clarification instead of fabricating information.

This honest approach builds trust.

Designed for Almost Every Specialty

Skriber works across a wide range of medical specialties. Internal medicine, psychiatry, cardiology, family practice, and many others use it effectively. The only limitations are specialties that require body diagrams or visual markings. For everything else, Skriber adapts to the doctor’s style.

If an engineer cannot configure a specialty template, the case goes all the way to Dr. Yost. He adjusts the template himself to make sure it reflects real clinical documentation.

Growing Integrations With Major EHRs

Skriber currently integrates with major platforms such as Epic, Athena, and eClinicalWorks, with more integrations being added. The depth of integration depends on access permissions at each hospital or clinic, but the goal is simple: place the note exactly where the physician needs it.

Conclusion: A Human-Centered Approach to AI Scribing

The interview made it clear that Skriber is not just another AI scribe trying to ride a trend. It is a tool built from the ground up with real clinician insight. The founders understood the burnout and frustration caused by documentation. Dr. Yost understood how AI really works — and how physicians think. Together, they created a system built around humility, accuracy, efficiency, and clinical reality.

Skriber succeeds because it respects the physician’s voice. It improves billing accuracy. It speeds up documentation. It removes friction. And through a simple but powerful habit — priming — it turns AI scribes from a novelty into a tool that truly saves time.

References:

• Healthcare IT Today. Priming the Ambient AI Medical Scribe with Skriber