My research interests include agentic AI, large language models, and machine unlearning. I am especially interested in unlearning in Stable Diffusion—developing methods to selectively forget specific concepts while maintaining model utility and generation quality.

Before graduate school, I spent over six years as a System Software Engineer on HP’s Imaging Team, specializing in image quality assessment, imaging pipeline optimization for scan applications, and on-device security solutions with OCR. This production experience—working with imaging pipelines and edge AI under real deployment constraints—now grounds both strands of my research: building trustworthy LLM-based agents and studying controllable behavior in generative image models such as Stable Diffusion.

That same systems-oriented mindset extends to how I work day to day. Through hands-on experience automating repetitive workflows in industry settings, I have seen how much reliable tooling matters for productivity. I aim to develop flexible agentic tools that meaningfully support research and academic work—tools that are dependable in practice, not just promising in prototype.