Happy July —
Before this month's update, a word on the whole idea. It comes down to three things — capture the moment, perform the work, preserve the judgment — across the least supported decade of a physician's career, from the first year of medical school to the first year of independent practice. We start with residency. Everything below is that idea turning into product.
The news I promised last month: the first cohort is committed and moving toward go-live. After a spring of demos and a Phase 1 waitlist that filled faster than I expected, the programs are now actively preparing with us for launch. I want to be precise about where things stand, because there is a meaningful difference between a committed pilot and a platform in production use—and I would rather describe that honestly than round up.
Here is the operating picture: July is dedicated to baseline measurement, configuration, kickoff, orientation, and the remaining readiness checks at each site. Procedure Logging and Evaluations—the two core workflows every program wants to try first—are built, staged, and ready to activate. Beginning in August, residents and Faculty will start using them in their day-to-day workflows as each program completes its readiness steps. This issue explains those two workflows and the runway from pilot preparation to go-live.
─────────────────────────────────
Where the Pilots Stand
When I wrote in June that Phase 1 was full, a fair question followed: full and doing what? So here is the real state of things, without the marketing gloss.
Every committed site is running on an existing system — New Innovations or MedHub — and none is doing a full swap-out on day one. That was always the model: run alongside, replace when ready. What is happening in July is the unglamorous, essential work that determines whether a pilot succeeds:
Baseline capture. Before we turn anything on, we measure the "before." How long does a procedure log take today? How long does an evaluation actually take, start to finish? What is the completion rate, and how late are they? You cannot claim to have saved time if you never measured the starting line.
Evaluation metrics defined. We are agreeing with each program, up front, on what success looks like — time-to-log, evaluation turnaround, completion rates, and honest qualitative signal from residents and faculty. Metrics set now, not reverse-engineered later to flatter the result.
Configuration and setup. Each site is moving through specialty, state, and class-list setup; persona provisioning across web and mobile; and activation of 13 curated clinical, educational, regulatory, and coding data sets. The pilot is designed without PHI and does not connect to the EHR, which materially simplifies setup and security review.
Kickoff and orientation. Short sessions with residents, faculty, coordinators, and Program Directors so the first live week feels familiar, not foreign.
Then, beginning in August, we activate the two core workflows described below together with the supporting pilot capabilities and begin measuring against the baseline. That sequencing is deliberate. A pilot that goes live before its metrics are defined is just a demo with extra steps.
─────────────────────────────────
Early Pilot Findings
Baseline input from participating program leaders points to a clear distinction between current workflow burden and future readiness. Satisfaction was very low across both procedure logging and evaluation workflows. Procedure logging was generally manageable for participating program leaders today, although this baseline did not include resident input. Leaders nevertheless identified limited visibility, inconsistent documentation detail, and variation in supervision and approval expectations. Evaluations present a more immediate program-staff burden, with delayed or incomplete forms, limited status visibility, and uneven evidence quality complicating CCC review and competency decisions.
The larger concern is future readiness. Participants expect competency-based medical education to add substantial work unless programs can capture faculty observations consistently, connect evidence to competencies and milestones, and prepare clearer support for CCC review and Program Director attestation. The opportunity is not simply faster tasks, but more reliable, visible, and useful evidence for residency decisions.
─────────────────────────────────
Turning On: Procedure Logging
Family Medicine residents at our pilot sites will be among the first to use a purpose-built, AI-assisted voice workflow for procedure logging. It closes the gap between the clinical moment and the documentation — which is exactly where logging usually goes to die.
A PGY-2 finishes a knee injection between patients. Instead of promising to log it later (and not), the resident shakes the phone, talks for ten seconds, and the AI does the rest.
Procedure Logging: shake, speak for ten seconds, confirm, submit
Behind that ten seconds, the AI drafts a high-quality log using 13 curated clinical, educational, regulatory, and coding data sets — including ABFM procedures, complications, ICD-10 diagnoses, ACGME milestones, ABFM competencies, duty-hour rules, and more. It fills the procedure, indicators, findings, complications, supervision, and codes. The resident confirms, makes any edits, and sends it for approval. Ten seconds of talking replaces several minutes of tapping through fields nobody enjoys.
─────────────────────────────────
Turning On: Evaluations
The second core workflow can turn an evaluation that once took fifteen minutes into a two-minute task — and, more importantly, make evaluations better. The problem it solves is not typing speed. It is memory. The multi-institutional median is now 217 faculty evaluations per trainee per year, most reconstructed weeks after the fact. You cannot write a good evaluation about a rotation you barely remember.
Evaluations are paired with observations captured by voice or text. Faculty retain each observation as private evidence. When completing an evaluation, they may select relevant observations and other approved evidence to support an AI-created draft. The faculty member reviews, edits, confirms, and submits the final evaluation.
Evaluations: capture an observation; select the evidence; review, edit, confirm, and submit the AI-created draft.
Faculty can capture a fresh observation in about fifteen seconds. At evaluation time, the AI creates a draft from the faculty member’s selected voice or text observations, other approved evidence, and the same 13 curated data sets — including the 14 ABFM competencies in use today (the 15th arrives in 2027). It can support the narrative across overall assessment, medical knowledge, patient care, interpersonal skills, professionalism, and comments. Faculty complete the ratings, review and adjust the narrative, confirm, and submit. The goal is not merely faster writing; it is a better evaluation grounded in observations captured in the moment rather than reconstructed from memory weeks later.
One important pilot requirement came directly from pilot feedback. Residents and Faculty should never lose a captured clinical moment simply because they are interrupted before completing the workflow. To support that, we are adding Moment Inbox™, a private workspace for incomplete procedure logs and faculty observations. When a capture is successfully preserved, the application confirms that it has been saved, and the user can return later to review, edit, delete, or finish it—without repeating the original voice capture. The promise is simple: once the application confirms a moment is saved, it should never have to be recreated from memory.
─────────────────────────────────
Why these two, first?
Procedure Logging and Evaluations were selected because they address two common, visible pain points, work naturally alongside MedHub or New Innovations, and can be measured clearly during the pilot. More importantly, they let us validate a core part of our vision: that a clinical moment captured once can carry the administrative work that builds from it — helping programs capture better evidence when it is created, reduce the effort of reconstructing it later, and make that evidence more timely, visible, and useful for residency decisions.
What Goes Live—and When
Here is what the next several weeks look like, so pilot programs know exactly what to expect and everyone else can see how we work. Each site begins with two core workflows—Procedure Logging and Evaluations—supported by Moment Inbox™ and the broader pilot foundation: Action Rail™, Insight Rail™, Chase Workflows, Morning Briefing, and tailored dashboards.
July — baseline and setup. Metrics are defined, systems are configured, teams are oriented, and readiness steps are completed. Features remain staged, not yet live.
Beginning in August — pilot go-lives. Each site begins using the two core workflows as its readiness steps are completed. Measurement begins against the July baseline.
Day 30 / 60 / 90 — honest feedback. Three short check-ins via a simple form. Not testimonials — just what worked, what did not, and what we should build that we are not.
Following the pilot — broader availability. Timing will be based on pilot readiness, product performance, and what we learn from residents, faculty, coordinators, and Program Directors.
The Phase 1 waitlist remains open, and I expect to add slots as pilot capacity becomes available. Later this year, we also expect to open pilots for the Onboarding, Scheduling, Reimbursement, and Recruiting Managers. The programs that raise their hands first will help shape what gets built.
─────────────────────────────────
Coming in August: National Resident and Faculty Surveys
Next month, GME Manager will launch national surveys of residents and residency faculty to better understand administrative burden, evaluation workflows, procedure logging, supervision documentation, CCC preparation, and perspectives on AI-enabled support in residency training.
We will share a summary of the national findings in September through webinars, conference presentations, and program visits. Programs and organizations that help distribute the surveys will also receive the more detailed findings, including the underlying de-identified survey results.
We are looking for volunteers to help us reach residents and faculty across specialties, program types, and regions. Program Directors, Program Coordinators, DIOs, faculty leaders, chief residents, and others willing to forward one or both surveys can make an important contribution to this national listening effort.
Interested in supporting this national listening tour? Contact Michael at [email protected], and we’ll send everything you need to share one or both surveys.
─────────────────────────────────
Interested in a Pilot?
If your program wants to be considered for the next available pilot opening, now is the time to raise your hand. The deal has not changed: pilots are free for 90 days, with no credit card, no commitment to convert, and no sales call at the end. Setup is fast and handled by us.
─────────────────────────────────
Know a Family Medicine PD who’d enjoy this?
Do me a favor and forward this. Tell them: “Someone finally did it: the work, captured in the moment it happens. One day we’ll tell residents and new faculty how we used to do it and they won’t believe us.”
Forward this newsletter →
With gratitude,
— Michael Sousa
Founder, GME Manager
[email protected] | 720-352-5203 | gmemanager.ai
─────────────────────────────────
Follow GME Manager on LinkedIn.
Pilot go-lives, early lessons, and the honest week-to-week of building GME Manager will be shared there. A follow and one share help more than you might think. Follow GME Manager →