About Me

Where it started

My path into ML began earlier and more playfully than you'd expect: in my Calculus I Honors course, I wrote a research paper on the math behind Snapchat-style facial filters – how derivatives and linear approximation drive the motion tracking behind facial keypoint detection. It was the first time I saw how the calculus I was learning connects directly to computer vision and ML, and it's what got me curious enough to start looking for research opportunities.

That curiosity led to my first internship at the National Institute of Standards and Technology (NIST), starting the summer after my freshman year at Montgomery College. The following fall, I picked up a second research thread: an NSF-funded CURM grant in combinatorial game theory, where our research group analyzed a Nim variant called Nearest Neighbor Nim – proving theorems about winning positions and building a calculator to compute game values that no existing tool supported. I presented this work at the Maryland Collegiate Honors Council Conference. We later turned the research into NNNim, a strategy game we piloted with middle schoolers through Sonya Kovalevsky Day, as a way to teach math through play.

What that's looked like at NIST

At NIST, I've been working on a machine learning project translating code between industrial robot languages – teaching models to bridge "dialects" that different robot brands speak, so factories don't need engineers to manually rewrite code for every brand. It was my first real exposure to how research-grade ML gets built end to end: from foundational math, through dataset construction (I wrote automation that cut data-generation time by roughly 180x), to fine-tuning transformer models and honestly assessing what didn't work yet.

Outside research

I've spent a lot of time in rooms where someone needed something explained simply: tutoring 30+ peers weekly in CS and math, teaching coding at after-school programs across three middle schools, and mentoring through Phi Theta Kappa and the Southern Management Leadership Program. I think the ability to explain something well is underrated as a technical skill, and it's one I actively practice.

Where I'm headed

This summer, I'm working on two projects. One is for my data science class, where I had the idea to focus on something in the "AI for good" space – building an NLP tool for mental health detection, with careful attention to the ethical considerations that come with that territory.

The second is more personal: a clinical NLP project for Wolof, aimed at making healthcare information more accessible to Wolof speakers. The idea came from following GalsenAI's work on WAXAL – Google's open-source speech dataset for 21 African languages, including Wolof – and seeing how community members stepped in to fix critical data issues so the Wolof subset could actually be used. That's the kind of work I want to be part of, and it's the project that feels most like my long-term direction.

When I'm not coding, I'm usually behind a camera or thinking about style – you'll find some of that on the Style & Expression page. I think creative practice and technical practice feed each other more than people give them credit for.