I am a system researcher by training and am particularly passionate about designing systems mechanisms and policies to handle trade-offs in cost, performance, and efficiency for emerging applications. Since joining WPI, my group’s research has focused on improving system support and performance for machine learning (ML)-related workloads. Because ML is widely adopted in many applications, it is critical to have performant systems that can effectively train, serve, and manage ML models. For example, a cost-effective cloud training framework can help ML practitioners innovate without requiring expensive on-premise GPU servers.

Specifically, we worked on providing mechanisms and policies for cloud-based distributed training, improving deep learning inference performance for mobile applications, and designing tailored optimizations for applications, including augmented reality, serverless computing, and video streaming. Our research is supported by National Science Foundation, Google Cloud, and VMWare Research. It has led to publications in top venues in the computer system (e.g., ICDCS, VLDB, MobiSys, SoCC) and machine learning  (e.g., ECCV, KDD, ICML).

Visit Digital WPI to view student projects and research advised by Prof. Guo

More news and media about Prof. Guo:

• Tian Guo Receives Outstanding Achievement Award | Worcester Polytechnic Institute

• Tian Guo Receives NSF Award for Research on Mobile Applications | Worcester Polytechnic Institute

• Video Interview with Prof. Guo