I recently graduated with a Ph.D. degree in Computer Science from the University of California, Riverside (UCR). I was supervised by Prof. Mohsen Lesani.
I’m interested in the theory and practice of distributed systems, computer security, and programming languages. I’m particularly interested in the foundation of consensus and distributed system protocols in the heterogeneous trust setting.
Currently, my research focus is designing efficient consensus protocols with open membership.
I’m currently a Research Scientist at Meta.
Sep/2024: I joined Meta as a Research Scientist.
Sep/2024: I successfully defended my Ph.D. dissertation “Heterogeneous Byzantine Replicated Systems”. Big thanks to all the people who have supported me through this journey. Special thanks to my cat Boba and dog Ori.
Aug/2024: Our work (Brief Announcement: Reconfigurable Heterogeneous Quorum Systems) is accepted to DISC’24.
Mar/2024: I completed my internship at Chainlink Labs. Thanks for all the support from my mentor Dr. Chrysa Stathakopoulou and my manager Lorenz Breidenbach. We worked on censorship-resistance reputation systems for leader-based BFT consensus.
Mar/2024: I am selected to receive Dissertation Year Completion Fellowship for 2024.
Feb/2024: I serve as a mentor for SIGPLAN Mentorship program, which is organized by Symposium on Principles of Programming Languages (POPL).
Sep/2023: I started my internship at Chainlink Labs as a research engineer. So excited to get hands-on experience in the industry research environment.
Sep/2023: I passed my dissertation proposal on “Heterogeneous Byzantine Replicated systems”.
Jul/2023: I am selected to receive the department scholarship to attend Grace Hopper Conference 2023.
Jul/2023: Our work (On the power of quorum subsumption for heterogeneous quorum systems) is accepted to DISC’23.
Mar/2023: I am selected to receive Dissertation Year Program Fellowship for 2023/2024.
Nov/2022: I am selected to receive student travel grant for 2022 ACM CCS.
May/2022: I am selected to receive GSA travel grant for 2022 IEEE Symposium on Security and Privacy.
Apr/2022: I am selected to receive student travel grant for 2022 IEEE Symposium on Security and Privacy.
Oct/2021: I am selected and funded to attend PLMW@SPLASH 2021.
Jul/2021: Our work Hamraz is accepted to IEEE S&P’22 (accept rate ~15%)!
Jul/2020: Welcome to my talk in CAV’20 Session 3B about our Hampa paper.
Jul/2020: Passed my Ph.D. candidacy exam!
Apr/2020: Our work Hampa is accepted to CAV’20 (44 papers accepted out of 241).
Sep/2019: Transferred to CS Ph.D. program at University of California, Riverside.
Sep/2017: Attended University of California, Riverside for CS MS program.
Jul/2017: Got my B.E. in Information Security from HUST(Huazhong University of Science and Technology) as OutStanding Graduates.
Reconfigurable Clustered Byzantine Replication: We design reconfiguration protocols for SMR across multiple clusters. We formally prove the safety and liveness properties of the reconfigurable clustered Byzantine replication and implement this framework for both HotStuff and BFT-Smart. Our experiment results show that clustered BFT SMR provides high throughput and low latency with dynamic membership.
Reconfigurable Heterogeneous Quorum Systems: We present reconfiguration protocols for heterogeneous quorum systems(HQS) in order to enable open membership for permissionless blockchains. We also presents trade-offs for the properties that reconfigurations can preserve, and accordingly, presents reconfiguration protocols and proves their correctness. We further present a graph characterization of HQS, and its application for reconfiguration optimization. I implemented reconfiguration protocols on top of Stellar blockchian. Our results are available on arXiv now: Open
On the power of quorum subsumption for heterogeneous quorum systems: We prove an impossibility result that shows quorum intersection, and quorum availability is not sufficient for Byzantine reliable broadcast(BRB) and consensus in heterogenous quorum systems(HQS), where each process can select different quorums. Moreover, we propose quorum-subsumption to help achieve BRB and consensus with detailed protocols and proofs. Quorum-subsumption together with intersection and availability is so far the weakest condition we know for BRB and consensus in the HQS setting.
Hamraz: How to have confidentiality, integrity, and availability with as few machine as possible for replicated objects? Better yet, we got the work done automatically. Our open-source project is available: Hamraz: code
Hampa: How to keep the invariant and recency of replicated objects as user specified with as little coordination as possible? Skip the manual labor and try out our tool with the help of program synthesis and the state-of-art SMT solver. Our tool and open source project are available: Hampa: tool, Hampa: code