About me

I am a graduating PhD student (2025 Summer) in Computer Science department at Columbia University, co-advised by Prof.Asaf Cidon and Prof.Junfeng Yang.

I received my B.S. from the University of Wisconsin-Madison in Dec 2019. At Madison, I was advised by Prof.Shivaram Venkataraman.

My research focuses in the intersection of Machine Learning, Systems and Security. I am actively exploring potential start-up ideas. Please contact me if you want to collaborate or just for a fun chat.

Recent News

• [present]. I am developing a testing framework for AI-agents focusing on completness, coverage and consistency. Stay tuned!
• [2024-11]. We conducted the first large-scale analysis on the usage of LLM among real-world malicious emails, partnered with Brracuda Networks! Results are coming soon!
• [2024-11]. We developed two robust AI-generated content detection systems on text & audio (tested on 1.3M samples, including real scam calls and simulations using tools like Elevenlabs)! Results are coming soon!
• [2024-11]. Talk @ AI & Cybersecurity Night hosted by the Google CISO Community. Thanks Google for the invitation!
• [2024-10]. Our work "Nazar: Monitoring and Adapting ML Models on Mobile Devices" on continious AI system monitoring is accepted by ASPLOS 25'!
• [2024-06]. I will start my summer internship at Barracuda Networks working on AI security in production!
• [2024-05]. Our work "MGit: A Model Versioning and Management System" on ML model lineage tracking and its applications is accepted by ICML 24'!
• [2023-05]. Selected as ML and Systems Rising Star 23'!
• [2023-03]. I will start my summer internship at MSR Redmond working on Project Fiddle again. In person this time!
• [2022-07]. Win the MLSys student travel grant!
• [2022-02]. I will start my summer internship at MSR Redmond. My mentors are Deepak Narayanan and Amar Phanishayee!
• [2022-01]. Our work "A Tale of Two Models: Constructing Evasive Attacks on Edge Models" is accepted at MLSys 22' (acceptance rate: 20.6%)! This is my first first-author paper during the PhD!
• [2020-11]. Clockwork has received the OSDI Distinguished Artifact Award! I wish they can give us a best video award as well!
• [2020-08]. Our work at MPI on DNN serving "Serving DNNs like Clockwork: Performance Predictability from the Bottom Up" is accepted at OSDI 20'!
• [2020-05]. I will start my summer internship under Prof.Jonathan Mace at MPI-SWS!
• [2020-04]. Paper accepted at ISBI 2020 with oral presentation!
• [2020-03]. I decided to goto Columbia for PhD!
• [2019-10]. Abstarct Paper accepted at AI Systems workshop at SOSP 2019 with lightning talk!

Vistor Map