May 8, 2026 Jialin Li — National University of Singapore
Distributed systems today face increasing challenges in performance, scalability, and cost. In this talk, I will present three works across different application domains that apply novel system-level designs to address these challenges. The first work, HyperEdge, significantly reduces the operational cost of a global-scale CDN without compromising video streaming quality. HyperEdge employs a pool of underutilized low-cost devices scattered across the globe, and carefully incorporates centralized management to guarantee content delivery performance. The second work, Cortex, designs a novel semantic caching layer to improve the performance and cost efficiency of LLM agents. The last work, Capybara, applies a software-hardware co-designed approach to implement a new load balancing architecture that addresses the limitations of existing L4 and L7 load balancers.
Jialin Li is the Sung Kah Kay Assistant Professor in the School of Computing at the National University of Singapore. He completed his PhD at the University of Washington in 2019. His research work has been awarded best paper awards at OSDI and NSDI. His current research interests include co-designing distributed systems with modern hardware, data plane operating systems, and distributed AI/ML systems.