Summary
Yuan Xia is a research-driven software engineer and PhD candidate in Computer Science at USC with nine years of hands-on experience in ML-driven program verification, autonomous systems, and LLM engineering. He has led research and engineering efforts at Nokia Bell Labs and USC, contributing novel methods to improve LLM reasoning accuracy, training/inference pipelines for Llama-family models, and generating concurrent invariants for distributed systems using SymbolicGPT. His work spans statistical verification with conformal inference for cyber-physical systems, collision-free multi-agent planning, and practical bug-hunting techniques for autonomous driving stacks, with peer-reviewed publications stemming from his undergraduate research. Comfortable moving between research prototypes and production-oriented ML tooling, he combines deep theoretical rigor (3.95 GPA in his PhD) with practical implementation skills across transformers, probabilistic models, and robotics simulators. Based in Los Angeles, Yuan pairs a poetic curiosity—“God speaks through Aerodynamics” resonates in his interdisciplinary approach—with a track record of patented innovations and industry internships at Microsoft.
9 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, 3.95, Doctor of Philosophy - PhD, Computer Science, 3.95 at University of Southern California
University of California, Irvine