Schedule: AI Research Symposium: Demystifying AI in Public Health

Everything You Wanted to Ask About AI (But Didn’t Know Whom to Ask)

A hands-on, plain-language day where researchers and practitioners unpack what AI can actually do for public health—right now and going forward. You’ll get quick primers and live demos that turn buzzwords into usable workflows, plus “ask-me-anything” moments throughout. Two interactive panels connect Large Language Models (LLMs), retrieval, and social listening to real decisions, then show how to turn risk estimates into actions that help people. Attendees will leave with a plain language understanding of what AI can (and cannot) do for public health today, plus fresh ideas and new places and partners to pursue them.

TIMEACTIVITIES
8:30 am– 9:00 am 

Breakfast in Diboll Gallery, 1440 Canal Street

 

9:00 am – 9:15 am
 

Opening Remarks: Dean Thomas LaVeist 

  • Welcome and Introduction
  • Overview of the symposium's goals and objectives 
9:15 am – 9:45 am 

15min – Samuel
15min - Patrick 

AI, Large Language Models, & Machine Learning in Public Health 
Speakers: Samuel Kakraba (WSPH), Patrick Button (CAIDS) 

  • Opportunities with AI/ML in public health
  • Overview of Tulane's current AI initiatives and future plans

9:45 am – 11:00 am 

Presentations 
10 min presentations x 5 


15 min moderated Q&A 
 

Speaker Session 1: Language Models, Social Listening & Retrieval-Augmented Tools in Public Health: NLP/LLMs/RAG and population-level text signals (social media, online narratives); “knowledge-to-guidance” systems grounded in curated sources. 
 

  • From LLMs to Agents - Harnessing AI to Unlock Insights from Biomedical Literature, Hua Xu, School of Medicine, Yale University
  • Evidence Synthesis Using Generative AI, Jagpreet Chhatwal, Harvard University/Massachusetts General Hospital
  • Retrieval-Augmented Generation (RAG) for Mindfulness: A Chatbot Grounded in an Evidence-Based Protocol, Simone Skeen, Warren Alpert Medical School, Brown University.  
  • Tracking public health signals from social media using natural language processing, Aron Culotta, Center for Community Engaged AI, Tulane University
  • Fast, Cheap, Rigorous? Lessons from AI and Social Listening in Public Health, Martha Silva, WSPH, Tulane University 

Audience Question & Answer  

Moderated by Paul Hutchinson
 

11:00 am – 11:15 am COFFEE BREAK & Quotes from Tulane Faculty AI survey 
11:15 amRaffle #1 Paul Hutchinson
11:15 am – 12:15 pm

Panel Discussion 1: Grounded Language Models and Public Signals 
Hua Xu, Jagpreet Chhatwal, Simone Skeen, Aron Culotta, Martha Silva  
 

Plain-language discussion of how AI and LLMs can support public-health work today, while also addressing common concerns for educators. Panelists will share near-term wins, common pitfalls, and simple practices. 

Moderated by Sam Kakraba

12:15 pm – 1:00 pm 

LUNCH & STUDENT AI POSTER SESSION

 Diboll Gallery and Grab-and-Geaux, Poster session in main first floor hallway

1:00 pmRaffle #2 Paul Hutchinson

1:00 pm – 2:15 pm

Presentations 
12 min presentations x 4 

(10 min ‘buffer’ built into transition between presentations, leave a little wiggle room for presentation overages)

Speaker Session 2: Clinical Prediction, Causal Inference & Equity with EHR and Population Data  
 

  • Utilizing big data without domain knowledge impacts public health decision-making, Miao Zhang, NYU Chunara Lab
  • Design and Implementation of Scalable and Equitable A/ML Workflows for Population-Health Measurement and Disease Diagnostics to Improve Public Health Outcomes in Low-Resource Settings, Samuel Kakraba, WSPH, Tulane University
  • Using modern risk engines and machine learning / artificial intelligence to predict diabetes complications: A focus on the BRAVO model, Lizheng Shi, WSPH, Tulane University
  • Computable phenotyping, machine learning, and trail emulation in the NIH RECOVER program. Tom Carton, Louisiana Public Health Institute (LPHI) 
     
    Audience Question & Answer

    Moderated by Patrick Button
2:15 pm – 3:00 pm

Panel Discussion 2: From Prediction to Impact: Making AI Work in the Real World 
Miao Zhang, Yilu Lin, Tom Carton, Hua Xu 
 

What happens after a model says, “high risk”? This panel explores how to move from risk scores to real-world action—checking whether interventions actually help, ensuring fairness across groups, and setting up data practices that teams can live with. Panelists will share the small, practical steps any team can take today.

Moderated by Lizheng Shi

3:00 pm – 3:15 pmCOFFEE BREAK 
3:15 pmRaffle #3 Paul Hutchinson

3:15 pm – 4:00 pm 

Presentation 
15min presentation

30 min moderated Q&A 

  • Getting things off the ground: Innovation & Funding
  • Tobin Lasson (Cedar Gate & DAC Member, Tulane WSPH)
  • Kimberly Gramm (Innovation Institute, Tulane)
  • Patrick Button (CAIDS, Tulane)

    A candid conversation with innovators and funders on how they choose, shape, and sustain AI projects.  
    Moderated by: Dean Thomas LaVeist 
    --- 
    3:15 – 3:30      Cedar Gate Technologies Presentation, Tobin Lassen 
    3:30 – 4:00      Panel discussion, moderated by Thomas LaVeist 
4:00 pm

Presentation of Poster Certificates & Wrap Up

Samuel Kakrab