Sujaya Maiyya

Assistant Professor

University of Waterloo

I am an assistant professor at Cheriton School of Computer Science at University of Waterloo. Prior to joining Waterloo, I did a short postdoc at Cornell University. I received my PhD from UC Santa Barbara where I was co-advised by Amr El Abbadi and Divy Agrawal.

Research Interests: My research interests broadly lie in distributed systems & databases and data privacy & security. I, along with my wonderful students, design, prototype and evaluate protocols for managing large-scale data in a performance efficient and secure manner. Currently, my group is solving prolems across different aspects of oblivious databases such as tunable-privacy, scalabily and fault tolerance, and support for relation queries. Please see more details in the Ongoing Research tab.

New Area of Research: Recently, I have been working on building platfoms to execute genomics algorithms privately using trusted execution environments (TEEs)! If you are interested to work on TEEs or in transforming insecure genomics algorithms to secure ones, please reach out to me!

Prospective Students: I am looking for highly-motivated PhD students with full funding opportunities. If you are interested in working with me on systems research, please send me an email and mention my name as a potential advisor in your application.

Recent News

  • December 2024: Thank you CFI and ORF for awarding our infrastructure grant proposal! News article
  • December 2024: I co-organized Ontario Database Day OnDBD-2024! Thanks to all participants for making the event a huge success!
  • November 2024: Was invited to speak at the CS-CAN DSDS workshop!
  • October 2024: Thank you NCC for awarding two research proposals - one on securely managing outsourced data and one on genomic data privacy! News article
  • April 2024 : Our paper, ORTOA, that develops a mechanism to hide the type of operation was accepted and presented at EDBT 2024!
  • September 2023: Gave an invited talk at University of Chicago, Northwestern, and University of Illinios Chicago on Waffle and QuORAM!
  • Aug 2023: Our paper, Waffle, that develops a tunably private database is accepted at SIGMOD 2024!
  • June 2023: Was invited to speak at the CS-CAN DSDS workshop!
  • April 2023: Thank you NSERC for awarding the Discovery Grant!

Education

PhD in Computer Science University of California, Santa Barbara (Jan 2018 - June 2022)
MSc in Computer Science University of California, Santa Barbara (Sept 2016 - Dec 2017)
BE in Information Science PESIT, Bangalore (Aug 2010 - May 2014)

Awards

  • CFI/ORF Infrastructure Grant 2024-2029.
  • NCC Research and Development Awards 2024-2028.
  • NSERC Discovery Grant 2023-2027.
  • MIT EECS Rising Stars 2021.
  • IBM PhD Fellowship 2020 award.
  • Google PhD Fellowship 2020 award (declined).
  • Outstanding Graduate Student 2018-19 award in the CS department at UCSB.
  • Elected Graduate Student Representative for 3 consecutive years at UCSB.
  • SAP Labs Scholarship (4 times) for securing the highest CGPA in the department at PESIT.
    •

    Publications

    Sujaya Maiyya, Sharath Vemula, Divyakant Agrawal, Amr El Abbadi, Florian Kerschbaum "Waffle: An Online Oblivious Datastore for Protecting Data Access Patterns" SIGMOD 2024 [pdf]

    Sujaya Maiyya, Yuval Steinhart, Adrian Davila, Jason Du, Prabhanjan Ananth, Divyakant Agrawal, Amr El Abbadi "ORTOA: A Family of One Round Trip Protocols For Operation-Type Obliviousness" EDBT 2024 [pdf].

    Mohammad Javad Amiri, Daniel Shu, Sujaya Maiyya, Divyakant Agrawal, Amr El Abbadi ". Ziziphus: Scalable Data Management Across Byzantine Edge Servers" ICDE 2023 [pdf].

    Sujaya Maiyya, Seif Ibrahim, Caitlin Scarberry, Divyakant Agrawal, Amr El Abbadi, Rachel Lin, Stefano Tessaro, Victor Zakhary "QuORAM: A Quorum-based Replicated ORAM Datastore" USENIX Security 2022 [pdf] [video].

    Sujaya Maiyya, Ishtiyaque Ahmad, Divyakant Agrawal, Amr El Abbadi "Samya: Geo-Distributed Data System for High Contention Data Aggregates" ICDE 2021 [pdf][slides].

    Fuheng Zhao, Sujaya Maiyya, Ryan Wiener, Divyakant Agrawal, Amr El Abbadi "KLL±: Approximate Quantile Sketches over Dynamic Datasets" VLDB 2021 [pdf].

    Sujaya Maiyya, Danny HB Cho, Divyakant Agrawal, Amr El Abbadi "Fides: Managing Data on Untrusted Infrastructure" ICDCS 2020 [pdf].

    Mohammad Javad Amiri, Sujaya Maiyya, Divyakant Agrawal, Amr El Abbadi "SeeMoRe: A Fault-Tolerant Protocol for Hybrid Cloud Environments" ICDE 2020 [pdf].

    Sujaya Maiyya, Faisal Nawab, Divyakant Agrawal, Amr El Abbadi "Unifying Consensus and Atomic Commitment for Effective Cloud Data Management" VLDB 2019 [pdf][ppt].

    Victor Zakhary, Mohammad Javad Amiri, Sujaya Maiyya, Divy Agrawal, Amr El Abbadi "Towards Global Asset Management in Blockchain Systems" BCDL co-located with VLDB 2019 (Demo) [pdf].

    Sujaya Maiyya, Victor Zakhary, Mohammad Javad Amiri, Divy Agrawal, Amr El Abbadi "Database and Distributed Computing Foundations of Blockchains" SIGMOD 2019 (Tutorial) [pdf].

    Sujaya Maiyya, Victor Zakhary, Divy Agrawal, Amr El Abbadi "Database and Distributed Computing Fundamentals for Scalable, Fault-tolerant, and Consistent Maintenance of Blockchains" VLDB 2018 (Tutorial) [pdf] [slides].

    Vaibhav Arora, RKS Babu, Sujaya Maiyya, Divyakant Agrawal, Amr El Abbadi, Xun Xue, Yanan Zhi and Jianfeng Zhu "Dynamic Timestamp Allocation for Reducing Transaction Aborts" IEEE Cloud 2018 [pdf].


    Ongoing Research

    Private Vector Database

    With the advent of large AI and language models, managing data in an embedding or a vectorized form is becoming increasingly common. Vector databases are customized DBMSs that effectively store and search embeddings. However, an embedding can reveal much about the underlying data, raising privacy concers. In this line of research, we are developing data systems that enable securely storing and searching on data embeddings.

    Oblivious Query Execution

    Hardware enclaves such as Intel SGX/TDX and AMD SEV enable applications outsource their data storage and processing needs to the cloud without compromising data confidentiality. However, side channel attacks on hardware enclaves effectively uncover the plaintext data process within the enclaves. We are building highly efficient and parallelizable primitives to process data securely in hardware enclaves without side channel leakages. Our primitives use vectorized query execution (this vectorization is different from representing data embeddings as a vector) and SIMD instructions to securely parallelize query execution.

    Building a scalable and fault-tolerant ORAM system

    Oblivious RAM or ORAM is a well known technique that preserves data access privacy by hiding access patterns. While there exists many variants of ORAM solutions, most are not scalable nor fault-tolerant - two vital features necessary for any practical data system. This project proposes the first scalable and fault-tolerant ORAM system that uses replication to tolerate bounded number of crash failures, while preserving privacy.


    Students

  • Nafis Ahmed (MMath - thesis) 2024-2026
  • Bishwajit Bhattacharjee (MMath - thesis) 2024-2026
  • Amin Setayesh (MMath - thesis) 2023-2025
  • Haseeb Ahmed (MMath - thesis) 2023-2025
  • Joseph Boulis (MMath - thesis) 2023-2025
  • Malav Mehta (Undergrad Researcher)
  • Mikaela Tang (Undergrad Researcher)
  • Aaryan Shroff (Undergrad Researcher)
  • Cheran Mahalingam (Undergrad Researcher)
  • Adrian Davila (Undergrad Researcher)
  • Jason Du (Undergrad Researcher)
  • Joline Cheng (Undergrad Researcher)
  • Emily Chen (Undergrad Researcher)
    •

    Teaching

  • CS348: Introduction to databases - Spring 2024
  • CS848: Privacy enhancing data systems - Winter 2024
  • CS348: Introduction to databases - Spring 2023
    •

    Professional Service

    Chairs
  • 2024: Ontario Database Day
  • 2023: SoCC Publicity Chair
    •  Program Committee
  • 2026: SIGMOD, EDBT
  • 2025: VLDB, EDBT
  • 2024: VLDB, ICDE Industry Track, CODS COMAD
  • 2023: ICDE, SoCC, SIGMOD Demo
  • Frequent reviewer: TKDE Jornal, DKE Journal, FGCS Journal