Building AI Agents for Healthcare: A Practical Introduction Using Microsoft Copilot Studio

Presentation Type: Workshop
Not recorded for AMIA Now

Presentation Time: 10:30 AM - 12:30 PM

Abstract Keywords: Generative AI in Clinical Workflow: Ambient Listening, Chart Summarization, Automated Response with LLM, Clinical Decision Support and Care Pathways, Education and Training, Analytical Artificial Intelligence: ML, Digital Pathology, Imaging AI, Predictive Analytics, Governance, Workforce Automation, Communication, and Workflow Efficiency, Ethics, Infrastructure and Cloud Computing, Innovation Partnerships, Implementation Science, and Learning Health Systems
Primary Track: Big Data for Health


Clinical AI agents powered by retrieval-augmented generation (RAG) can transform how clinicians access institutional knowledge—from guidelines and protocols to operational procedures. This focused workshop provides a practical introduction to building clinical AI agents using the DKITE Framework (Define, Knowledge, Integrations, Test & Telemetry, Evolve). Using a Microsoft Copilot Studio sandbox environment, participants explore the spectrum of AI agents—declarative, task-based, and autonomous—while learning foundational concepts applicable across AI platforms. The session covers essential RAG architecture including vectorization, embeddings, semantic search, and knowledge base design for structured and unstructured data sources. Through hands-on exercises, participants analyze clinical scenarios including pathology and lab intelligence agents, clinical trial matching, and ambulatory scheduling optimization. Participants configure a basic knowledge source and build a prototype agent. The workshop emphasizes practical evaluation using a five-dimension rubric and introduces adoption analytics for monitoring agent performance. Advanced patterns including vision-based GUI agents for legacy systems are discussed. Participants leave with a workflow analysis template, condensed DKITE Framework reference, and an evaluation checklist for immediate application in their organizations.

Speaker(s):
Mikael Guzman Karlsson, MD, PhD, FAAP
Baylor College of Medicine | Texas Children's Hospital

Matthew Bayes, MD, MPH
Baylor COM / Texas Children's Hospital

Lauren Hess, MD, FAAP, ABPM-CI
Baylor College of Medicine

Thomas Chong, MD, MS
Texas Children's Hospital/ Baylor College of Medicine

Author(s):
Mikael Guzman Karlsson, MD, PhD, FAAP - Baylor College of Medicine | Texas Children's Hospital; Matthew Bayes, MD, MPH - Baylor COM / Texas Children's Hospital; Lauren Hess, MD, FAAP, ABPM-CI - Baylor College of Medicine; Thomas Chong, MD, MS - Texas Children's Hospital/ Baylor College of Medicine;

---