I do research, experimentation of machine learning systems for Medical devices development at Medtronic Surgical.
My academic research focus on the reliability of medical vision-language models.
I believe safe and educated adoption of AI-driven systems can make healthcare equitable around the world.
My beliefs
We can help create a world we’re glad to live in.
Hover or tap to read more.
Build things that are useful to people.
I care about where the work ends up. In healthcare, I want what I build to help someone make a better decision or make good care more accessible. That purpose helps me choose which problems are worth spending time on, even when the answer is a long way off.
Understand why something works.
A good result makes me curious about what produced it. I want to understand the assumptions behind it and the conditions under which it fails. Finding a failure can be useful progress. It gives us something specific to investigate and a better question to ask next.
Make problems easy to talk about.
I want to hear about a concern while there is still time to act, even if we cannot fully explain it yet. My response should make it easier for people to speak up again. I try to thank them for raising it, then help work out what we need to do.
Give people a clear goal and room to act.
I want people to understand why the work matters and who is responsible for moving it forward. That responsibility should come with the authority to make decisions. I explain the constraints, offer honest feedback, and trust the person doing the work to choose how to approach it.
Help people grow through the work.
I want people to take on work they could not have done a year earlier. My part is to give them a chance, stay available when they need help, and let them develop their own judgment. When the work succeeds, the people who did it should get the credit.
Stay open to being wrong.
I try to ask what would change my mind, especially when I like my own answer. When the evidence points elsewhere, I need to follow it and explain what I got wrong. If I was confident in public, I should be willing to correct myself there too.





















