This abstract was part of the 2026 UCSF AI and Education Symposium
The rapid expansion of generative AI tools in medical education has created both opportunity and risk. While large language models can accelerate content creation and learner support, most educational implementations remain generic, poorly aligned with curricular standards, and disconnected from real learner needs and clinical workflows. Lusitropy was developed to address this gap by operationalizing the concept of precision education—delivering educational content that is structured, contextualized, adaptive, and accountable to evidence-based frameworks.
Lusitropy is an AI-enabled educational platform designed to support trainees rotating through pediatric cardiac anesthesia, a domain characterized by high cognitive load, heterogeneous learner backgrounds, and limited structured learning time. Rather than positioning AI as a replacement for educators, Lusitropy embeds AI within a deliberately constrained educational architecture. Content is organized around clinically meaningful topics and cases, aligned to learning objectives, and delivered through modular pathways that integrate reading, case preparation, adaptive tutoring, and formative assessment.
This show-and-tell will demonstrate how Lusitropy integrates multiple AI-driven components into a coherent learning ecosystem. Attendees will see how structured educational content is generated and curated using prompt scaffolding and predefined schemas; how learner interactions are captured to assess engagement, progression, and areas of difficulty; and how AI tutoring is used to support synthesis and reflection rather than rote answer generation. The platform also incorporates question-based practice, case-based workflows, and analytics that allow educators to examine learner behavior without increasing supervisory burden.
A central focus of this presentation is demonstrating how Lusitropy augments the clinical work flow and delivers AI generated content safely. Lusitropy is designed to make AI behavior observable and auditable, emphasizing automated prompt design, faculty response evaluation, and alignment with educational intent. The platform demonstrates how educators can move beyond ad-hoc AI use toward systems that respect curricular goals, learner variability, and institutional expectations for responsible AI use.
During the session, the presenter will walk through a real use case: a trainee preparing for a pediatric cardiac case using Lusitropy’s guided workflow. The demonstration will highlight how AI is used to scaffold preparation, identify knowledge gaps, and support reflective learning, while maintaining educator oversight and control of content boundaries and pedagogical priorities.
This show-and-tell is intended for educators, trainees, and AI leaders interested in moving from experimental AI adoption to sustainable, principled implementation. By sharing both design decisions and lessons learned, Lusitropy offers a concrete example of how AI can be harnessed to enhance—not dilute—medical education in complex clinical domains.
Contact
Rishi Kadakia, [email protected]