Yu Gan
Yu Gan

Senior Research Engineer

About Me

Yu Gan is a senior research engineer at Google AI and System Research, working on efficient LLM serving and coding agents. He received his Ph.D. degree in Electrical and Computer Engineering at Computer Systems Laboratory (CSL), Cornell University, advised by Professor Christina Delimitrou. His Ph.D. thesis was focused on building and managing large-scale microservices in datacenters, improving their performance and resource efficiency using machine learning. He is currently broadly interested in ML for systems and systems for ML. Before studying at Cornell, he obtained my B.Eng. degree in Electronic Engineering from Tsinghua University in 2016.

Yu Gan has been recived three IEEE Micro TopPicks Awards and one CIDR Best Paper Award.

Interests
  • Efficient LLM serving
  • LLM agents for coding
  • ML for systems
  • Cloud computing
Education
  • Ph.D. Electrical and Computer Engineering

    Cornell University

  • M.S. Electrical and Computer Engineering

    Cornell University

  • B.Eng. Electronic Engineering

    Tsinghua University

📚 My Research

I am a senior research engineer at Google AI and System Research. I am generally interested in the interdisciplinary area between machine learning and computer systems. My research focuses on three main pillars:

  • Systems for ML: where I design scalable and efficient infrastructure, combining the insights into ML algorithms, to support ever-growing model complexity and request load;
  • ML for Systems: where I leverage machine learning to automate and optimize the design and management of computer systems.
  • LLM Coding Agents: Building and exploring the capabilities of autonomous LLM-based agents for software development, debugging, and optimization.

Ultimately, my work aims to create a virtuous cycle where advancements in AI and systems mutually propel each other forward.

Please reach out to collaborate 😃

Featured Publications
Recent Publications
(2025). EchoLM: Accelerating LLM Serving with Real-time Knowledge Distillation.
(2025). Towards Foundation Database Models. In CIDR 2025.
(2024). CHASE-SQL: Multi-Path Reasoning and Preference Optimized Candidate Selection in Text-to-SQL. In ICLR 2025.
(2024). Sleuth: A Trace-Based Root Cause Analysis System for Large-Scale Microservices with Graph Neural Networks. In ACM ASPLOS'23.
(2023). Ditto: End-to-End Application Cloning for Networked Cloud Services. In ACM ASPLOS'23.

Experience

  1. Senior Research Engineer

    Google AI and System Research

    Research projects include:

    • Efficient LLM serving via online knowledge distillation
    • Building text to SQL agents with self-consistency and reinforcement learning
    • Cardinality prediction benchmarks and models with GNN and graph transformers
    • Foundational models for table and quary representation
  2. Senior Software Engineer

    Alibaba Cloud

    Research projects include:

    • Building trace analysis tool for large scale microservices with GNN

Education

  1. Ph.D. Electrical and Computer Engineering

    Cornell University
    Thesis on Designing and Managing Large-Scale Interactive Microservices in Datacenters. Supervised by Prof. Christina Delimitrou. Presented papers at 5 IEEE conferences with the contributions being published in 2 Springer journals.
    Read Thesis
  2. M.S. Electrical and Computer Engineering

    Cornell University
    Courses: Machine Learning, Statistical Distances in ML, Computer Architecture, Operating Systems, Advanced Systems, Cloud Computing
  3. B.Eng. Electronic Engineering

    Tsinghua University
    GPA: 91.3/100