AI is changing what we can build. My direction is to put that capability at the service of care, connecting science, software and people.
These are convictions that guide my work and hypotheses I want to investigate. I revisit them as I learn from research and practice.
01 THESIS
AIshouldgivetimebacktocare.
I believe in tools that extend physicians’ capabilities and reduce the work that takes them away from patients. Technology needs to fit the daily reality of care.
My starting point is the problem experienced by clinicians and teams. Automation makes sense when it improves their workflow and keeps human responsibility in decisions.
IN PRACTICE
Design with the people who will use it, test on a small scale and observe time, effort and quality before expanding.
02 THESIS
Reliabledatacomesbeforeextraordinarymodels.
My thesis is that data quality, context and validation will be decisive in turning models into useful healthcare tools.
A promising result on one dataset does not guarantee performance in another service or population. I want to build systems whose limitations can be understood and evaluated.
IN PRACTICE
Document data provenance and quality, evaluate in the intended setting and monitor failures after deployment.
03 THESIS
Ophthalmologycanopenpathsformedicine.
I see ophthalmology as a field to investigate, test and demonstrate applications of AI in healthcare. It is where I focus my research and build experience.
Images, signals and clinical questions create opportunities in computer vision and oculomics. Each application needs its own evidence; extending it to other fields remains a hypothesis to investigate.
IN PRACTICE
Start with a specific clinical question and bring statistics, engineering and healthcare professionals together.
04 THESIS
Researchneedstomeetpractice.
I want to bring scientific discovery, products and real-world use closer together. An idea gains value when people can test, understand and use it.
Publishing and deploying are different steps. My direction is to connect scientific rigor with software that can be maintained and evaluated in daily work.
IN PRACTICE
Turn questions into experiments, experiments into prototypes and verifiable results into product decisions.
SCIENCE · SOFTWARE · PEOPLE
This is where I choose to study, build and share.
Explore my research and professional journey, or get in touch to exchange ideas.