researchOngoing research · public conference abstracts
What a model can carry from one setting to another
Collaborative surgical-AI research on generalizability, annotation, and the limits of a promising result.
Surgical AI Research Lead · co-author
Andres Bravo
I build humane systems for high-stakes health.
My work connects surgical AI, public-health analytics, and biomedical innovation. I focus on what a result means in its setting, what remains uncertain, and how it can support a useful next step.
Good research should make the next decision clearer.
- Clinical care
- Responsible AI
- Research
- Public health
- Product innovation
- Shared learning
Static view
Between a result and a useful next step
Good research should make the next decision clearer.
A strong model score, a stable cluster, and a useful tool answer different questions. I want to understand how the evidence was produced, what setting it describes, and what someone could reasonably do with it.
Start with the decision
Before choosing a model or building a prototype, I ask what someone needs to decide. The task, the setting, and the people affected should shape the technical work from the start.
Show the reasoning
A score or a pattern needs an explanation. I want the assumptions, comparisons, and uncertainty to be clear enough that another person can trace the reasoning and challenge the conclusion.
Keep judgment visible
A technical output should help someone reason through a problem. Reviewers need to know what a system can support, where it stops, and when they should revise, reject, or pause its use.
Selected work
The questions behind the work
- public healthOngoing internship · public methods articles
From patterns to careful interpretation
Responsible youth-persona work at WDGPH, described through public methods and safeguards.
AI4PH Data and Analytics Intern · author
- researchPublished narrative review
Compare like with like
A collaborative narrative review of transformer-based surgical AI.
Co-author
- innovationEducational concept and prototype
A clearer next step, with a human boundary
A patient-navigation concept and prototype exploring clearer next steps in care.
Concept and prototype developer
How I want to work
Be precise
Name the task, the evidence, and the limit of the claim.
Protect dignity
Consider how a label, workflow, or disclosure could affect the person behind it.
Make decisions reviewable
Document choices so that others can understand them and disagree.
Share the work
Credit collaborators and make room for knowledge outside my own discipline.
About Andres Bravo
A life beyond the work
My cultural roots are Cuban, Italian, and Canadian. Language and belonging are part of the context I bring to work across disciplines and communities.
Outside work, I make room for running, cycling, strength training, golf, and recreational soccer. I also enjoy sketching, with earlier time spent painting and sculpting.
AboutWhere I want to take this work
I am working toward becoming a physician-innovator. I want to bring this practice closer to care: understanding the clinical question, examining the evidence, and building tools that support the person making the decision.
What are you trying to understand?
If you are working on a question in health AI, public health, or biomedical innovation, I would be glad to compare approaches and learn what you are seeing.