About Me
I'm Yoshi, a student researcher investigating how confidential computing and hardware-based trust can reshape cloud infrastructures.
Currently a junior majoring in Computer Science & Communications Engineering at Waseda University, where I focus on secure systems research. My work lies at the intersection of confidential computing, trusted execution environments, and distributed systems.
Research Affiliations
Kasahara-Kimura Laboratory
Green Computing Systems Center • Waseda University
Securing cloud workloads while making them more useable and flexible with IO devices.
IIJ Research Laboratory
In collaboration with Imperial College London's LSDS Group
Applying serverless workloads to confidential computing infrastructures for next-generation cloud security.
News
-
August-September 2025
Joining Prof. Fan Zhang's Decentralized Systems Group at Yale University for a summer research internship, researching anonymity networks and TEEs!
-
June 21, 2025
First preprint uploaded for my WiP paper on Confidential GPU Computing:
-
June 2025
Volunteering at ISCA 2025, held this year at Waseda University.
-
June 2025
Hosting a GenAI Workshop at Google Japan (Shibuya) through Kuma Lab, a student-led research community I co-founded.
Research
In a world where data is the new currency, trust and privacy are no longer optional—they're foundational.
How do we build cloud infrastructures that are not just fast and scalable, but fundamentally trustworthy?
At the intersection of hardware, software, and virtualization security lies a uniquely complex trust problem. My research addresses this through confidential computing and its real-world integration.
Current Research Projects
Serverless Confidential Containers (SC2)
Researching Kubernetes serverless deployments on confidential computing infrastructure to enable privacy-preserving serverless cloud. This project bridges the gap between modern cloud-native architectures and hardware-based security guarantees.
Technologies: K8s, Knative, Confidential Containers (CoCo), Kata Container, Intel TDX, AMD SEV-SNP
GPU Throughput Evaluations for CVMs
ActiveMeasuring GPU throughput overheads across different CVMs to understand I/O performance over different vendor hardware. This research aims to quantify the performance trade-offs when running GPU-accelerated workloads in confidential environments.
Research Interests
Foundation of Trustworthy Cloud Computing
TEE-based anonymity networks & serverless
Bioinformatics & federated learning
Contact
Feel free to reach out if you have questions about my research, are interested in collaboration, or want to connect.
yoshisato@kasahara.cs.waseda.ac.jp
Green Computing Systems Research Center (Bldg. No. 40)
School of Fundamental Science and Engineering, Waseda University, Shinjuku, Tokyo, Japan