I build systems that have to work under real conditions, not ideal ones, embedded software that reads noisy sensor data mid-flight and neural networks that give real results. I'm less interested in what looks finished and more interested in what's actually load-bearing underneath it. If a system can't survive contact with reality, it isn't done.

My direction is clear: research-grade work at the intersection of statistics and machine learning, the kind of problems that don't have clean answers yet. I'd rather spend years learning the system than build something simple. The foundations I'm building now are meant to hold weight for a long time.

My standards don't shift based on visibility. The rigor I bring to a project no one is evaluating is the same rigor I bring to one that is. Consistency is the true measure of discipline. I approach problems by working back to their root cause rather than settling for a surface-level fix, and I don't consider something finished until I understand why it works, not simply that it does.