CV
General Information
| Full Name | Mingyuan MA |
| Date of Birth | 30th April 2001 |
| Languages | Chinese, English |
Education
-
2023 - 2025
Master of Science in Data Science
Harvard University, MA
- Cross-Registration in EECS at MIT
-
2019 - 2023
Bachelor of Arts (Double Majors)
University of California, Berkeley, CA
- Computer Science & Statistics, High Distinction Honors
Professional Experience
-
Jul 2025 - Present
Software Engineer, LLM Inference Workload Performance
NVIDIA, Santa Clara, CA
- End-to-end LLM inference performance benchmarking, analysis, and automation infrastructure development
-
May 2024 - Aug 2024
Software Intern, LLM Inference Workload Performance
NVIDIA, Santa Clara, CA
- LLM inference performance analysis platform with multi-agent system for interactive querying
-
Jun 2023 - Aug 2023
Deep Learning Algorithm Engineer Intern
Moonshot AI (Kimi), Beijing, China
- Efficient LLM architecture with tokenization-free parallel decoding and adaptive token merging
Research Experience
-
Oct 2024 - May 2025
Research Collaborator
SGLang | Sky Computing Lab, UC Berkeley
- Distributed GPU-sharing inference system for multi-LLM serving with 3.3x SLO and 2x cost reduction
-
Sep 2023 - Sep 2024
Research Assistant
Microsoft Research Lab - Asia (MSRA)
- Mutual reasoning framework with MCTS for enhancing Small Language Models' reasoning capability
-
May 2022 - Jul 2023
Research Assistant
HPC-AI Lab, National University of Singapore
- Continual learning of vision-language models with zero-shot transfer preservation