Kris Young
π Kexin Chu
Ph.D. Student in Computer Science
University of Connecticut | Focusing on Machine Learning Systems, LLM Infrastructure, System Security, and Disaggregated Memory
π Connecticut, USA
π‘ Open to Collaboration! If you're interested in MLSys, Multi-Agent Systems, Disaggregated Memory, or Security, feel free to reach out β let's build something exciting together!
π§ͺ Research Interests
π€ LLM Systems
Efficient inference, KV-cache optimization, and serving systems for large language models
π System Security
Security and privacy in ML systems, timing side-channel mitigation
πΎ Memory Systems
RDMA, disaggregated memory, CXL, and memory-tiered architectures
π Selected Publications
MCaM: Efficient LLM Inference with Multi-tier KV Cache Management
ICDCS 2025 Conference PaperMulti-tier KV cache management system for efficient large language model inference.
π Read Paper βExpertFlow: Adaptive Expert Scheduling and Memory Coordination for Efficient MoE Inference
arXiv MoE SystemsRuntime system for MoE inference that combines adaptive expert prefetching and cache-aware routing to optimize inference under memory constraints.
π Read Paper βSelective KV-Cache Sharing to Mitigate Timing Side-Channels in LLM Inference
Security arXivSecurity-focused approach to prevent timing attacks in LLM serving systems.
π Read Paper βeInfer: Unlocking Fine-Grained Tracing for Distributed LLM Inference with eBPF
eBPF 2025 WorkshopeBPF-based tracing framework for distributed LLM inference systems.
π Read Paper βSafeKV: Safe KV-Cache Sharing in LLM Serving
MLArchSys 2025 ISCA 2025 SecurityPrivacy-preserving KV-cache sharing mechanism for multi-tenant LLM serving.
π Paper π₯ Presentationπ Education
Ph.D. in Computer Science
University of Connecticut, USA | 2024 - Present
Research Focus: ML Systems, KV-cache optimization, RDMA-backed storage, and disaggregated memory architectures.
π° Predoctoral Fellowship Recipient
M.S. in Integrated Circuit Engineering
Hefei University of Technology, China
Co-supervised by A.P. Ying Wang and A.P. Cheng Liu
Specialized in computer architecture and AI acceleration
B.S. in Integrated Circuit Design & Integrated System
Hefei University of Technology, China
Foundation in digital circuit design, computer organization, and system integration
π National Scholarship Recipient (2018, 2019)
πΌ Industry Experience
Software Architect & Backend Engineer
Baidu Inc., Beijing, China
Department: Search R&D Platform - Focus on large-scale backend systems
π Career Progression: T3 β T4 (2021) β T5 (2023)
Key Contributions:
- DeepQA Web Services: Developed high-performance web services using C++/brpc for large-scale search infrastructure, handling millions of queries daily
- LLM Access Control Systems: Built comprehensive access control and monitoring systems for Ernie Bot/WenXinYiYan (ζεΏδΈθ¨) using Golang, Redis, and MySQL. Ensured security and reliability for production LLM services
- Real-Time Streaming Systems: Designed and implemented Kafka-based streaming pipelines for large-scale data indexing and ingestion, enabling real-time data processing
- System Architecture: Led architectural design decisions for high-throughput, low-latency distributed systems serving billions of requests
π Honors & Awards
π Predoctoral Fellowship
University of Connecticut
2025
π Baidu Pride Special Award
Baidu Inc.
2022
π National Scholarship
China Ministry of Education
2018, 2019
π οΈ Technical Skills
Programming Languages
ML/AI Frameworks & Tools
Systems & Infrastructure
Databases & Message Queues
Research Areas
π¬ Get In Touch
Interested in collaboration or have questions about my research?
Send me an email β