📲 Add this app to your home screen for quick access
UM
AI Research Skills
University of Miami · Miller SOM
AI-Assisted Research Skills Workshops
University of Miami · Miller School of Medicine
Program at a Glance
…
Attendances
…
Participants
…
Sessions
About This Workshop
The AI-Assisted Research Skills Workshop
series at the University of Miami Miller School of Medicine teaches researchers,
faculty, and staff to leverage AI tools for data analysis, systematic reviews,
and educational app development.
Led by Dr. Lina Shehadeh,
Division of Cardiology & Interdisciplinary Stem Cell Institute.
🙏 With Gratitude
This program is made possible through the generous support of:
Want these hands-on AI workshops delivered to your faculty, department, or institution? See our tracks, objectives, and how to invite Dr. Shehadeh and her team.
All Sessions
Loading…
🔥 Community at a Glance
Loading insights…
📊 Who Attends?
Loading…
👔 Faculty Leadership & Departments
Loading…
🏛️ Faculty by Department
🩺 Inside Department of Medicine
Sub-breakdown of the Department of Medicine faculty across official UM divisions.
👔 Leadership Role Breakdown
🗓️ Sessions by Track
Loading…
💼 LinkedIn Highlights
Posts by Dr. Lina Shehadeh — tap any card to view on LinkedIn.
Materials are hosted on the University of Miami SharePoint site.
Sign in with your UM credentials (CaneID) when prompted.
For Institutions & Departments
Bring the AI-Assisted Research Skills Workshops to your institution
Hands-on, build-something training that leaves your faculty and trainees with working AI research skills — delivered in person by Dr. Lina Shehadeh and her team.
In-person onlyTrain-the-trainersCohort-based
🌱 Our Philosophy — Build the Builders
Hands on keyboards, not just slides
Every session, participants build a real artifact — a coded analysis, a screening pipeline, a deployed app — not a demo they watch.
Train the trainers
We equip your own faculty to keep teaching after we leave, so the capability stays inside your institution.
Real research, real tools
Python, AI-assisted screening, PRISMA-compliant workflows, and agentic app-building — mapped directly to the research tasks your people already do.
Why in person only? The hands-on, side-by-side cohort format is what makes the skills actually stick — so we deliver these on-site, not over video.
📚 What We Offer
Three tracks, each deliverable on its own. Jump to a track or send a direct link:
The entry point for everyone. No coding background required — participants use AI as a coding assistant to go from zero to a working data analysis in a single sitting. (This is also Session 1 of Track 2.)
👥 Faculty, trainees, staff🧩 No prerequisites💻 Laptops + Jupyter
You'll build: a health-data visualization, an AI-prompted Excel analysis, and a one-page financial dashboard.
Foundations · Python fundamentals · data viz · financial intelligence
01
Set up a working Python environment (Jupyter) and use AI as a coding assistant within a clear, ethical framework for research.
02
Grasp core Python fundamentals — data types, variables, and loops — and why Python matters for research workflows.
03
Build a guided data visualization from a real health dataset (e.g., smoking-risk data).
04
Use AI prompts to perform data analysis on spreadsheet/Excel data with confidence.
05
Independently build a one-page financial dashboard from data, applying the session's skills end-to-end.
06
Leave with reusable code, prompt patterns, and the confidence to apply AI-assisted coding to your own research.
Track 2 · Bootcamp Series
AI-Assisted Systematic Reviews
5 sessions
An end-to-end, PRISMA-compliant systematic-review pipeline — from research question to evidence tables — with AI accelerating every stage. Session 1 is the Track 1 foundation; Sessions 2–5 build the review.
Distinguish between systematic reviews and scoping reviews, including their methodological differences.
02
Formulate your research question using the PICO framework.
03
Understand the PRISMA and PRISMA-ScR frameworks for transparent reporting.
04
Search major databases, improve searches with AI, and export citations.
05
Work with RIS, BibTeX, and CSV citation formats.
06
Build a foundation for AI-assisted screening pipelines in subsequent sessions.
Session 3 — Deduplication Strategies
Preparing your dataset for AI-assisted screening
01
Understand why deduplication is critical for PRISMA-compliant systematic reviews.
02
Use EndNote for efficient, validated deduplication (practical method).
03
Apply Python-based deduplication for large-scale or reproducible workflows.
04
Implement exact matching (DOI/PMID) and fuzzy matching techniques.
05
Use AI/LLM to resolve ambiguous duplicate candidates.
06
Create audit trails and generate PRISMA-compliant documentation.
Session 4 — Title/Abstract Screening with AI
Building your AI-assisted screening pipeline
01
Transform your PICO criteria into effective AI screening prompts.
02
Build a Python script for batch title/abstract screening.
03
Generate screening decision logs for PRISMA documentation.
04
Implement confidence scoring and uncertainty handling.
05
Simulate dual-reviewer workflows with AI for validation.
06
Create quality-control checkpoints and validation sampling.
Session 5 — Full-Text Screening & Data Extraction
From PDFs to structured evidence
01
Organize and manage full-text PDFs for systematic review screening.
02
Extract text from PDFs using Python libraries.
03
Build AI prompts for full-text eligibility assessment.
04
Document exclusion reasons with specific citations from text.
05
Design and implement structured data extraction forms.
06
Create comprehensive evidence tables from included studies.
Track 3 · Intensive
Building Educational Apps with AI Agents
1 session · ~3 hrs
Participants describe a goal and let an AI agent plan, write, run, fix, and deploy a working educational app — building and showcasing a real interactive learning tool in a single session.
You'll build: a deployed cardiac action-potential learning platform — with a quiz and a clinical case — live.
Session 1 — Building an Educational App Using AI Agents
From "what is an agent" to a deployed, showcased app
01
Explain what an AI agent is and how it differs from a chatbot or an AI coding tool — it sets its own plan, writes/runs/fixes code, and deploys; you just describe the goal.
02
Use an AI-agent platform (e.g., Replit) to turn a plain-language description into a working educational app.
03
Iteratively extend an interactive learning tool — adding a quiz and a clinical case — on a real biomedical topic.
04
Apply agentic workflows to polish, debug, and deploy a functioning app within the session.
05
Showcase a deployed app and adapt the approach to your own teaching or research.
06
Leave with a reusable template and workflow for building AI-agent-powered educational tools — no prior software engineering required.
📈 Proven at Scale
48
Sessions
931
Attendances
370
Participants
23
Departments
UM Miller School of MedicineMeharry Medical CollegeSociety of Vascular & Interventional Neurology (SVIN)
🗓️ Formats & Logistics
Single foundational session
Track 1 — one ~3-hour hands-on session for a department or cohort.
Multi-session bootcamp
Track 2 — the full 5-session systematic-review series, scheduled to your calendar.
App-building intensive
Track 3 — one ~3-hour session where each participant deploys an app.
You provide: a room, participant laptops, and Wi-Fi. We bring: the curriculum, materials, live instruction, and a train-the-trainers handoff. In-person delivery only.
🤝 Invite Dr. Shehadeh & Team
Bring hands-on AI research training to your faculty and trainees. Tell us your audience, track(s) of interest, and timeframe — we'll design the delivery around your institution.
Dr. Lina Shehadeh, PhD, FAHA — Professor of Medicine, Division of Cardiology & Interdisciplinary Stem Cell Institute, University of Miami Miller School of Medicine. Founder, Empower AI Leaders. MIT Certificate in Data Mining & Machine Learning.
