I am a final-year Ph.D. candidate in Computer Science at The University of Texas at Austin, advised by Aditya Akella and Sanjay Shakkottai. I build system abstractions, frameworks, and generalizable AI components, grounded in theoretical analysis, that let real-world systems efficiently approximate optimal behavior as workloads, environments, and objectives change.
My work spans networking, operating systems, and machine learning. I was part of the founding proposal team for the NSF Expedition on Learning-Directed Operating Systems (LDOS), where I now lead the undergraduate research program. Before UT, I received my B.S. in Electrical and Computer Engineering from Shanghai Jiao Tong University. I have also spent time at Microsoft Research, Google, and Intel.
Research
Latent information should be a first-class object in system design.
Today's systems expose isolated measurements, delayed feedback, and the effects of only the deployed policy. The information that matters most for good decisions stays hidden. My research recovers this latent structure with lightweight AI, exposes it through reusable system interfaces, and uses theory to characterize what it can and cannot buy.
Representing latent workload and system dynamics
Build compact learned representations that recover workload and system state that raw measurements hide.
Guiding policy search with system knowledge
Exploit system knowledge to guide near-optimal policy search with smaller search space.
Characterizing limits and safeguarding
Theory for when extra information helps, and runtime guardrails that keep learned policies safe.
Recent News
- Oct. 2026I am on the 2026–27 academic job market, seeking tenure-track faculty positions in computer systems and networking.
- Sep. 2026Our paper Online Batching with Lookahead: Adversarial Limits and Stochastic Gains is accepted to SIGMETRICS 2027.
- Sep. 2026I am hosting the LDOS Undergraduate Research information session, introducing LDOS research and opportunities for undergraduates. If you are interested, please apply here.
- Jul. 2026My amazing undergraduate mentees Justin Chang and Angela Zhang are presenting their work on decentralized edge network scheduling at DEMO workshop at ICML 2026.
- May 2026I am presenting UNUM at NSDI 2026. Please check our paper for more information.
Selected Publications
-
SIGMETRICS '27Online Batching with Lookahead: Adversarial Limits and Stochastic Gains
-
NSDI '26
-
HotOS '25How I Learned to Stop Worrying and Love Learned OS Policies
-
SIGCOMM '23