I built an AI skill that reviews research manuscripts the way I review them.

When a trainee sends me a draft, they need tracked changes in the Word doc, margin comments explaining why, an email that tells them what to prioritize, and a rationale document I can reference later.

I built a skill that gives me a first pass.

It scans my emails for the research related items, downloads the manuscript and it reads the full paper section by section, then returns:

→ The original document with real Word tracked changes and comment bubbles

— the actual file they sent me, marked up → Methodological flags a peer reviewer would raise

→ Policy connections to active rulemaking (TEAM, No Surprises Act, price transparency) where the authors missed the thread

→ Suggested language for new paragraphs, not just "consider expanding"

→ A student-facing email grouping the feedback by priority

→ A rationale report documenting the reasoning behind every annotation

Think of it like having a grad student who's learned your preferences and does the initial markup before you sit down with it. I still read every comment, edit the language, and make the final calls.

The difference is that the closed loop with my trainees went from days to hours — which means I can mentor more students and stay involved in more projects without the bottleneck being my Sunday evening.

The first paper through the pipeline was a health economics manuscript on hospital ownership and amputation pricing. Five tracked changes, twelve margin comments, a journal targeting recommendation — all grounded in orthopedic surgery and health policy expertise.

I know people have strong feelings about AI in research writing and review. I think the question worth asking is whether the tool is replacing judgment or extending reach. For me, this is augmentation. It handles the structured first pass so I can spend more time on the mentorship and thinking that only a human advisor can do. But I welcome the conversation. If you see this differently, I think the conversation is a worthwhile one!

This is part of the Agentic Clinic — over 50 scheduled tasks, prompts, and downloadable skills for clinicians, researchers, and healthcare operators who want to work at a different speed. Peer review, clinical coding, patient journey design, meeting prep, deal pipeline tracking. All built by a surgeon who uses these tools every day. I'm rolling out these resources for Techy Surgeon Founding Members now.

If you're spending Sunday nights redlining student papers, this one's for you.

The Agentic Clinic is open! Join us Saturday Live on Techy Surgeon for learning how to build skills on your own and subscribe for discounted access to agenti-clinic.com

#HealthcareAI#AgenticAI#ClinicalResearch#MedEd#TheTechySurgeon