>

AI Deep Dives

AI Deep Dives (Most Fridays from 1pm – 2pm)

AI Deep Dives are a great time to explore research problems which might be solvable with deep learning. Discuss your ideas with Vanderbilt data scientists, other researchers, DS graduate students, and undergraduate students. These are hour-long sessions where together, we brainstorm and formulate solutions, lay out project plans, recruit students and engage potential collaborators for the goal of initiating new projects, new AI project phases, or new AI project directions. Bring your lab groups, fellow startup founders, and fellow colleagues, and together, we’ll pioneer a path towards your solution. Request an AI Deep Dive here!

Upcoming AI Deep Dives

  • February 6th: Exploring Voice AI and Mobile Solutions for Rural Heart Failure Patients

    Dr. Deonni Stolldorf (PhD, RN, FAAN) from Vanderbilt School of Nursing will lead a discussion on leveraging AI technologies to deliver the GUIDED-HF telehealth intervention to rural heart failure patients via mobile-accessible modalities.

    What we'll cover: We'll explore whether Audio OpenAI (voice AI) could effectively deliver a self-care intervention for heart failure patients, and examine alternative approaches including smartphone apps and chatbots. Key considerations include: designing for patients with phone-only internet access, reducing technical barriers to telehealth participation, and integrating daily symptom tracking. The session will brainstorm practical architectures and identify the most promising path forward for rural patient populations.

  • February 27th: AI-Powered Curricular Intelligence for Medical Education
    Shane Stenner from VUMC's Department of Biomedical Informatics will present his project developing an AI-powered curricular intelligence platform that transforms how medical students learn, faculty teach, and institutions govern medical education. This session will focus on the technical architecture and implementation strategy for grounding large language models in Vanderbilt's complete first-year curriculum.
     
    What we'll cover: We'll explore several interconnected challenges: how to chunk and embed multimodal content (slides, transcribed audio, PDFs) to optimize cross-course retrieval; learner modeling approaches that balance personalization with privacy; evaluation design for both internal decisions and external grant applications; adoption strategies for integrating with existing workflows (VSTAR, Learn, email); content quality guardrails and hallucination risk management; and sustainability considerations including LLM inference costs at scale and pathways to M2/clerkship expansion.
  • March 13th: Defining "Great" AI Output for B2B Marketing
    Brian Moyer from StrongPosition, a Nashville-based AI-powered B2B marketing startup, will lead a collaborative session exploring how to define and measure quality in AI-generated marketing content when traditional metrics don't apply.
     
    What we'll cover: StrongPosition helps B2B marketers keep positioning and messaging current in fast-moving markets. However, marketing output is inherently subjective—how do you distinguish "good" from "great" when there's no clear metric? This session will explore methodologies for defining quality standards for AI-generated content, discuss frameworks for expert validation, and consider how to build feedback loops that improve prompt engineering over time. We'll draw on founder John Farkas's decade of proprietary marketing methodology to ground the conversation in real-world effectiveness.
  • March 20th: AI-Enhanced Medical Education
    Dr. Bill Cutrer from Vanderbilt University Medical Center's Department of Pediatrics will lead a discussion on integrating AI tools into the MD curriculum. As the faculty member overseeing the medical school curriculum, Dr. Cutrer is exploring ways to leverage AI to enhance both student learning outcomes and faculty teaching effectiveness.
     
    What we'll cover: This session will explore the current landscape and emerging opportunities for AI integration in medical education. We'll discuss practical applications of AI tools that can support student learning—such as intelligent tutoring systems, adaptive assessments, and study aids—as well as tools that can help faculty design curriculum, provide feedback, and personalize instruction. The conversation will address implementation challenges, pedagogical considerations, and opportunities for collaboration across disciplines.

Past AI Deep Recaps