Hi! I'm Kristen, an ML Performance Engineer at Amazon Web Services, optimizing large language model inference on custom ML accelerators. I have 6+ years of experience spanning ML systems, performance optimization, and distributed systems, and am proficient in distributed inference, LLM internals, and kernel-level optimizations.

Beyond performance engineering, I'm deeply interested in how intelligent systems work; how architecture and computational constraints shape what models learn to represent, and what those representations reveal about the nature of intelligence itself.

I draw on cognitive science and philosophy of mind to inform how I think about the systems I build and study, and I write about these ideas here.