My work sits at the intersection of technical robustness and normative philosophy. I believe that as we build more powerful autonomous systems, the primary engineering challenge is no longer just performance, but value alignment.

Currently, I am investigating how we can embed deontological constraints directly into gradient-based optimization processes without sacrificing the generalization capabilities of neural networks.

When not at the lab, I advocate for inclusive AI policy and contribute to open-source governance toolkits that help smaller organizations audit their model biases.