My work focuses on translating technical complexity into clear, effective solutions across emerging and specialized domains. Recent projects include designing novel frameworks for time series forecasting, applying genetic reinforcement learning algorithms for reasoning in large language models, and developing fast methods for representation learning.

Prior to my work in computer science, I conducted research in mathematics with a particular focus on the Langlands program. My mathematical work spanned both the Galois and Lie theory perspectives. In Galois theory, I computationally verified Buhler’s thesis on the modularity of icosahedral Galois representations and investigated potential analogues for even representations. On the Lie theory side, my research centered on cohomological parabolic induction, with an emphasis on infinite-dimensional representations of real reductive Lie groups and automorphic representations of $GL(n)$.

While my focus has since shifted toward computer science, mathematics remains a central influence on my work. The rigor and structural depth of the field provide a foundation that informs my approach to contemporary problems in machine learning and beyond.