It's 4:47 PM. You've seen 14 patients. Your ambient scribe nailed every note: SOAP structure, ICD codes, the works.

And yet here you are. Still sitting at your desk.

Because the note was never the real bottleneck. The bottleneck is everything that comes after the note: the referral that needs routing, the four lab orders that need placing, the follow-up message your patient is waiting on, the care plan update you told yourself you'd get to "after clinic." The supplement protocol adjustment you mentally flagged during the visit, but haven't documented. The Oura data your patient uploaded last night still needs to be cross-referenced with this morning's labs.

Your scribe listened. It listened beautifully. But it didn't do anything.

If you're running a functional, integrative, or longevity practice where every patient generates three times the data and five times the workflow complexity of a conventional visit, you've probably felt this gap more acutely than most. The note is maybe 20% of your post-visit cognitive load. The other 80% is coordination, communication, and execution. And until now, that 80% has been entirely on you.

That's starting to change. And it's changing fast enough that it's worth your attention this week.

Over the past two years, ambient AI scribes have proven their value. The largest study to date just published in JAMA, tracked 8,581 ambulatory clinicians across five academic medical centers, including 1,809 who adopted AI scribes. The findings: 16 fewer minutes of documentation time per eight-hour shift, with primary care physicians, advanced practice providers, and women clinicians seeing the most pronounced gains. Clinicians who used the scribe for at least half their visits saw even bigger reductions — over 27 fewer minutes of documentation time per shift.

AI scribes work. If you haven't adopted one yet, that's still your single biggest unlock.

But if you have adopted one, you've probably noticed the limitation: the scribe's job ends when the note is signed. Everything downstream — the orders, the messages, the care plan updates, the team coordination — still falls to you.

A new class of AI tools is crossing that line. The industry calls them "AI agents" or "agentic AI," but the label matters less than the capability shift: these are tools that don't just capture what happened in a visit — they act on it. They chain together multi-step tasks across your systems. They follow protocols you've set. They execute the workflows you'd otherwise do manually after every single encounter.

The simplest way to think about it: a scribe is a transcriptionist. An agent is a well-trained practice coordinator who also happens to have perfect recall — someone you've told your preferences, your protocols, and your standards once, and who takes the note and does the next five things without being asked.

One important clarification before we go further: this is not autonomous clinical decision-making. The Stanford-Harvard State of Clinical AI report, released just this week, synthesizes the most meaningful evidence from 2025 and confirms what you already know intuitively — AI performs best when it assists rather than replaces the clinician. Agentic AI handles the execution layer. You set the protocols. You make the decisions. You approve the outputs. The agent handles the clicks, the routing, and the follow-through.

The question isn't whether this shift is coming. It's already here, in early form, at health systems and practices across the country. The question is what it looks like in your specific clinical reality — and what you can do about it right now.

The easiest way to understand what agentic AI changes is to see it inside workflows you already live in. Here are three.

A new patient shows up with five years of outside labs, a GI MAP, a DUTCH hormone panel, Oura ring data, a 12-page intake form, and a chief complaint of "I just want to know what's actually going on."

Most clinicians reading this are somewhere in the middle of a progression that's moving faster than it was six months ago. Here's a quick way to orient yourself:

You're still typing or dictating notes manually. Your biggest unlock right now is an ambient scribe — not agentic AI. Start there. The cognitive relief alone is worth it, and the JAMA data backs that up. We've covered this in previous issues, and we'll link the best resources at the bottom.

You have a scribe, but everything after the note is still on you. This is where most of you are. You're getting the documentation benefit, but the post-visit workflow — orders, messages, care plan updates, team coordination — is still manual. This is exactly the gap that agentic AI is designed to close, and it's the focus of this issue.

Your AI already handles some post-visit tasks. Maybe it drafts patient messages, or queues lab orders, or pre-charts before visits. You're in early-adopter territory. The next question for you is whether you can set persistent protocols — your preferences, your clinical logic, your workflow standards — so the system learns and improves over time instead of starting from scratch every session.

Your AI chains multi-step workflows across systems with minimal intervention. You're living in the future, and there aren't many of you yet. If this is you, we want to hear about it — reply and tell us what's working. We'd love to feature your setup.

If you're in functional, integrative, or longevity medicine, the tools that truly match your workflow complexity are still catching up. Most agentic AI platforms are built for conventional primary care — straightforward chief complaints, standard-of-care pathways, insurance-coded billing. Your world is different: multi-modal data streams, layered protocols, root-cause treatment logic that doesn't map neatly to conventional guidelines. Platforms like Vibrant Practice are being built specifically for this space. Knowing that gap exists helps you shop smarter and set realistic expectations.

An ambient scribe captures information. An agent acts on it. That's a meaningful jump in capability — and it means the due diligence is different. Before you let any AI tool move beyond listening and into doing, ask three things:

Every agentic tool should have a clear, visible boundary between "draft and wait for my review" and "execute automatically." If a vendor can't show you exactly where the human-approval checkpoints are, that's a dealbreaker.

The specifics matter here. Can the tool place a lab order, or only draft one for your signature? Can it send a patient message, or only queue it for your review? Can it update a care plan, or only suggest changes? There's a world of difference between an agent that prepares your work and one that executes it. You need to know which mode you're operating in at every step.

This is the single biggest gap for functional and integrative clinicians evaluating agentic tools. If an agent's clinical logic is built on conventional guidelines and standard-of-care pathways, it will flag your protocols as outliers. It might suggest a statin when you're managing cardiovascular risk through metabolic optimization. It might question a supplement protocol that doesn't appear in its formulary. It might route a "normal" lab result that you know is suboptimal based on functional ranges.

Ask whether the knowledge base is customizable. Ask whether it can learn your specific protocols, your preferred ranges, and your treatment logic. Ask what happens when your approach diverges from what the model considers standard. If the answer is "it follows conventional guidelines only," then you'll spend more time correcting the agent than it saves you — and that defeats the purpose.

Every AI system will produce errors. The question is how the tool handles that reality. Can you see the agent's reasoning — why it drafted a particular order, flagged a particular value, composed a particular message? Or is it a black box that gives you an output with no audit trail? If overriding the agent is harder than doing the task yourself, the tool isn't serving you — it's creating a new kind of administrative burden dressed up as efficiency.

It's 4:47 PM again. You've seen 14 patients. But this time, the post-visit work isn't waiting for you — because you built a system that handles the execution so you can stay in the part of medicine you trained for.

That's the shift. Not replacing you. Not automating your judgment. Just closing the gap between what you decide and what actually gets done — so the 80% of your day that's coordination and follow-through stops stealing from the 20% that's clinical thinking.

The three questions above are your filter. Use them to evaluate every tool that crosses your desk — whether it's a new EHR feature, a standalone agent, or a vendor pitch. If the tool can't tell you what it does without your approval, whether it understands your kind of medicine, and what happens when it's wrong, keep moving.

And if you're ready to stop evaluating and start building — Vibrant was designed for exactly this moment. It's an AI-native EHR built specifically for functional, integrative, and longevity practices: the workflows you actually run, the data streams you actually use, the clinical logic that conventional platforms don't support. Not a conventional system with AI bolted on. A platform built from the ground up for the medicine you practice.

Be a modern clinician with the help of Vibrant, the AI-powered EHR built specifically for personalized medicine.

One of the first randomized controlled trials of ambient AI scribes just published — a pragmatic three-arm trial of 238 outpatient physicians across 14 specialties, comparing two commercial scribes (Microsoft DAX Copilot and Nabla) against usual care. The results were more nuanced than the headlines suggest: Nabla users saw a meaningful 9.5% reduction in documentation time, while DAX showed no significant change compared to the control group. Both tools showed potential improvements in burnout and cognitive task load scores, though the researchers note those secondary findings need confirmation in larger, multicenter trials. The takeaway for clinicians: not all scribes are created equal, and the specifics of how a tool is built — and how consistently your team actually uses it — matter as much as whether you adopt one. If y

The inaugural State of Clinical AI report from the ARISE network — a Stanford-Harvard research collaboration — is the most comprehensive synthesis of where clinical AI meaningfully advanced in 2025 and where the gaps remain. The finding that should anchor your thinking: prediction tasks (deterioration risk, disease trajectory, risk scoring) show the strongest evidence, and AI consistently performs better as a second opinion layered onto clinical judgment than as a standalone decision-maker. If you read one thing this month to calibrate your AI expectations, make it this.

Claude Cowork: An AI Agent That Can Use Your Computer. Anthropic just expanded Claude's "computer use" capability to both Mac and Windows — giving Pro and Max subscribers a desktop AI agent that can open apps, navigate browsers, fill in spreadsheets, and execute multi-step workflows while you step away. It works through a feature called Dispatch: you assign Claude a task from your phone or desktop, and it uses a continuous screenshot-analyze-act loop to get it done. Think of it as the first real consumer-grade agentic AI tool. It won't write directly into your EHR yet, but for research synthesis, drafting patient communications, care plan documentation, and operational tasks, it's a practical preview of what's coming. The adoption numbers are notable too — Anthropic reports stronger early uptake for Cowork than Claude Code saw in its first comparable period. If you've been curious about what "agentic AI" actually feels like to use, this is the most accessible on-ramp right now.

Lilly Bets $2.75 Billion on AI-Discovered Drugs and Longevity Is the Target: Eli Lilly just signed a deal worth up to $2.75 billion with Insilico Medicine, an AI drug discovery company whose platform was built around aging biology. Insilico's target discovery engine, PandaOmics, identifies dual-purpose targets linked to the hallmarks of aging, and their molecular design engine generates novel compounds in days rather than years. They have 28 drugs in their pipeline, nearly half already in clinical trials, and their first compound went from target identification to Phase I in under 30 months — compared to the traditional 4-to-6-year average. For longevity clinicians, this is the pipeline story worth watching: AI-accelerated therapeutics specifically designed around the biology you're already treating. The compounds that emerge from this pipeline could reshape the protocols you're building in the next two to three years.

The calendar is filling up. A few worth putting on your radar - and a chance to connect with the Vibrant community in person!

Apr 8–11 — HINT Summit · Nashville, TN The anchor conference for Direct Primary Care and membership medicine. If you're evaluating or scaling a membership model, this is where the operational knowledge lives. Vibrant will be there!

Apr 13 — Vibrant Community Webinar “All Things Peptides with Leonard Pastrana, nuAgeBio“ Virtual Event Join Sunita and Leonard Pastrana to discuss the latest changes in the peptide regulatory landscape and what it means for your practice.

Apr 10–12 — A4M Spring Congress · West Palm Beach, FL The A4M's flagship spring gathering for anti-aging and longevity medicine practitioners. Strong mix of clinical science and practice building. Vibrant (just Sunita) will be there!

May TBD — Vibrant Community Webinar “AI Skills for the Modern Clinician“ Virtual. Join Dr. G and Dr. Sunjya Schweig to walk through the new competencies for the AI-savvy clinician, along with tangible tips for using a variety of tools in your practice.

May 27–30 — IFM Annual International Conference · San Diego, CA The largest gathering of functional medicine clinicians in the world. This year's program reflects the field's rapid expansion into AI-assisted care and multi-omics diagnostics. Vibrant will be there!

Jun 9–11 — Longevity Docs Cannes 2026 · Cannes, France Awards and summit for longevity medicine's emerging leaders. A smaller, high-signal event for clinicians building practices at the frontier.

Nov 5-7 — Private Physicians Alliance Annual Meeting · St. Petersburg, FL The gathering for independent, cash-pay, and concierge physicians navigating practice independence. Practical and peer-driven. Vibrant will be there!

Nov 8-11 — American College of Lifestyle Medicine Conference · Orlando, FL Lifestyle medicine's main annual event — evidence-based approaches to behavior change, chronic disease, and healthspan. Growing overlap with the longevity medicine community. Vibrant will be there!

TBA — Longevity Clinics Roundtables · Buck Institute Clinical practice meets research infrastructure. One for the clinicians who want to be closer to where the science is actually being made.

Know of an event we should add to the list? Reply and let us know.

The AI scribe was a real unlock. But if you're still spending your evenings on orders, messages, and care plan updates, you've only solved the first 20% of the problem.

The clinicians who will get the most out of this next wave aren't the ones who adopt fastest; they're the ones who ask the right questions before they trust a tool to act on their behalf. The three-question filter in this issue is a good place to start.

If you're already using agentic AI in your practice or you've evaluated a tool and walked away, reply and tell us what you're seeing. The best insights in this newsletter come from clinicians living it.

The goal was never to work faster. It was to spend your time on the medicine. We're getting closer.

—Sunita and Dr. G

The Modern Clinician is written for functional, integrative, and longevity-focused physicians who want to scale their impact and deliver cutting-edge care.

If you liked this one, share it with a colleague! We appreciate you spreading the word.

To learn more about the why behind this newsletter, start with our first post introducing The Modern Clinician.