Keduse Worku

Hi, I’m Keduse, a final-year PhD student in astrophysics at Johns Hopkins University finishing in Summer 2026. I work on building computational and statistical models to understand complex systems under uncertainty, and I’m transitioning fully into Applied Scientist and Data Scientist roles in industry.

Most of my work involves probabilistic modeling, large-scale simulation, and analyzing noisy, high-dimensional data where assumptions and validation really matter. During my PhD, I’ve built GPU-accelerated simulation pipelines in JAX, developed uncertainty-aware classification systems for large datasets, and worked extensively with Bayesian inference and time-series modeling.

I’ve also applied these skills in industry. Most recently, I worked in asset and liability modeling at New York Life, where I implemented callable bond pricing models under the Hull–White framework and developed numerical methods to support scalable fixed-income pricing workflows.

Alongside technical work, I care deeply about communication and collaboration. I enjoy translating complex models into clear, actionable insights, whether through research, cross-disciplinary work, or public science outreach.