About

I build bridges between theoretical machine learning and messy reality.

My background in biometrics and research taught me that the hardest part of AI isn't the model architecture—it's the data infrastructure, the deployment constraints, and the user trust. I care deeply about building systems that are robust, explainable, and actually solve the problem they were designed for.

Currently, I'm exploring how large language models can be effectively grounded in enterprise data without hallucinating, and building tools to make that process transparent.