I am a third-year PhD student in Computer Science at George Mason University (GMU), advised by Prof. ThanhVu Nguyen. My research focuses on the reliability of AI and optimization software: verifying deep neural networks (DNNs) and testing the solvers that verification tools depend on.

Research highlights

• Testing MILP solvers. Built a metamorphic testing framework with an exact, certificate-backed oracle for Gurobi, SCIP, HiGHS, CBC, and GLPK. Across about 85,000 solver runs it found 36 defects, 26 of them previously unknown; developers have confirmed 4 and fixed 4 so far.
• Bit-flip fault verification. Developed two GPU-accelerated verifiers for quantized DNNs under hardware bit-flip faults. OrBit solves 2.4× more benchmarks than the state-of-the-art tool, and mBFV verifies networks with up to 1.15M parameters under multiple simultaneous weight flips.
• NeuralSAT. Contributor to NeuralSAT, a DPLL(T)-based DNN verifier ranked 2nd in the International Verification of Neural Networks Competition (VNN-COMP) in 2024 and 2025.

I am also a Graduate Teaching Assistant for CS 310 (Data Structures) at GMU.

Before GMU

I received my B.S. in Computer Science from the University of California, Davis (UC Davis) in 2023. During my undergraduate studies, I contributed to educational and community projects including rvcodec.js under the supervision of Professor Porquet, Profficient at CodeLab, and Colorflix & TaskCollab at Google Developer Student Club @ UC Davis.

After graduation, I worked as a software developer at the Ronald Lab (UC Davis), where I built and maintained KitBase, an open-access rice genomics platform described in a 2026 article in Database, and developed pipelines for rice mutation analysis.