About
I am a Ph.D. student and Research Assistant at Nanyang Technological University (CCDS), advised by Prof. XiaoFeng Wang. My long-term goal is to strengthen runtime defenses for Confidential Computing without degrading confidentiality or performance. To that end, I study Programming Languages and Formal Verification and tie them to Operating Systems and AI Security for end-to-end designs.
Previously, I completed my master's at Carnegie Mellon University (INI) and collaborated with the Amazon AWS Kani team on model checking for Rust's standard library. As an undergraduate at Purdue XINU Lab, I integrated mouse/keyboard drivers into Xinu-x86-GUI, wiring the init path, IRQ handling, and scheduling/memory interactions to deliver stable GUI I/O for teaching.
Education
- Ph.D. in Computer Science, Nanyang Technological University, 2026–present
- M.S. in Information Technology — Information Security, Carnegie Mellon University
- B.S. in Computer Science, Purdue University
Ongoing Projects
Verified Information-Flow Control for Confidential Computing
Making security guarantees verifiable end-to-end for workloads running inside confidential virtual machines — even when the guest OS itself cannot be trusted. The goal: provable protection of sensitive data flows with acceptable performance overhead.
Practical Privacy-Preserving LLM Inference
Bridging the gap between the cryptographic promise of two-party computation and the infrastructure-scale demands of large language model inference — studying where the real performance costs live and how to make private inference deployable at data-center scale.
AI Security of Agentic Systems
Understanding new security risks that emerge when LLM-powered agents are deployed on real platforms: how attacks can persist and spread across agent ecosystems, and what defenses hold up in practice.
Agent-Assisted Protocol Security Analysis
Security flaws often originate in protocol specifications, not just implementations. We study how AI agents can help systematically surface these specification-level risks and validate their impact in real-world implementations.
News
- 2026.08 — Joined NTU CCDS as a Ph.D. student, advised by Prof. XiaoFeng Wang.
- 2025 — Started as a Research Assistant at Indiana University Bloomington.
- 2024 — Completed CMU × AWS Kani practicum; key NonZero proofs landed in CI.
- 2023 — Delivered Xinu-x86-GUI input drivers and regression tests at Purdue XINU Lab.