About Me
Here is Siming Deng (Larry Deng).
I am a graduate student in the Department of Electrical and Systems Engineering (ESE) at the Penn Engineering of the University of Pennsylvania. Currently, I build agentic AI tools and infrastructure for iPhone hardware testing, spanning agent runtimes (multi-agent orchestration and tool use), on-device inference with Apple MLX, closed-loop test optimization, and Vision Transformers for distribution diagnosis. Broadly, my interests lie at the intersection of AI infrastructure, applied ML, and hardware systems. Prior to joining Penn, I gained valuable research experiences as a Research Assistant at Shenzhen University advised by Dr. Jun Zhang. My work focused on Nonparametric Statistics, Distortion Measurement Error Models, and Large Sample Theory. I developed and analyzed advanced statistical methods, conducted large-scale simulations, and applied these techniques to real-world datasets, resulting in five SCI publications.
Work Experience
If you are interested in any aspect of me, I am always open to discussions and collaborations. Feel free to reach out to me at - siming_deng_stat [at] 163.com
Seeking Software Engineer and Machine Learning roles — AI Infrastructure, Applied AI / Agents, and ML Systems. Feel free to reach out!
Publications
Research Interests
- ML Systems & AI Infrastructure
- Efficient & On-Device Inference (Apple MLX, SGLang
) - Agentic AI (multi-agent orchestration & tool use)
- Machine Learning
- Nonparametric Statistics
I currently work at the intersection of AI infrastructure and applied ML, building agent runtimes and optimizing LLM inference.
Previously, my research centered on nonparametric statistics and measurement-error modeling, unobservable-variable estimation under multiplicative distortion frameworks, where I developed calibration methods and simulation studies to improve correlation-coefficient estimation in complex industrial settings, resulting in five SCI publications.
News and Updates
Apple unveils iPhone 18 Pro & iPhone Duo. Proud to have contributed to the agentic hardware-testing tooling behind the launch.
First-Author Paper: Innovative covariance-based framework: symmetry assessment and exponentiality testing under multiplicative distortion measurement Errors Now Officially Published in Communications in Statistics - Simulation and Computation
Excited to have received an offer from Apple!
Thrilled to have received an offer from UPenn Engineering!
First-Author Paper: A New Logarithmic Multiplicative Distortion for Correlation Analysis Now Officially Published in Statistical Analysis and Data Mining (JCR Q1)