Sujaya Maiyya
UC Santa Barbara
sujaya_maiyya@ucsb.edu
Bio
Sujaya Maiyya is a PhD candidate at UC Santa Barbara co-advised by Amr El Abbadi and Divy Agrawal. Her research interests lie in the intersection of data management distributed systems and data security. Sujaya is a recipient of IBM PhD Fellowship and Google PhD Fellowship (declined) awards for 2021-22. Sujaya is also recognized as an Outstanding Graduate Student (2019) for her leadership as a student representative and her commitment to the Computer Science department at UCSB. She has spent her summers solving problems in distributed systems in companies such as Google and IBM Research. Service and outreach programs to uplift underprivileged populations is a cause important to Sujaya and she volunteers at School on Wheels a non-profit organization that tutors students living in shelters group homes or without homes.
Managing data in both trusted and untrusted infrastructures
Managing data in both trusted and untrusted infrastructures
Individuals and enterprises produce over 2.5 exabytes (10^18 bytes) of data everyday. Much of this data – including sensitive and private information – is stored with and managed by third parties such as Amazon Web Services or Google Cloud. These companies can lose millions to billions of dollars in sales if their data access latencies increase by only a few hundred milliseconds. However there is a fundamental tradeoff between security and efficiency in data management systems. While these companies own and trust their underlying data infrastructure and so can prioritize latency the same is not true for those utilizing their services. My PhD research focuses on designing prototyping and evaluating data management protocols that strike a balance between efficiency and security in both trusted and untrusted systems. When the infrastructure is trusted and with efficiency as the focus I proposed two protocols: (i) to execute transactions on geo-distributed data that is both sharded (or partitioned) for scalability and replicated for fault-tolerance. Our protocol compared to GoogleÂ’s database Spanner commits 27-88% more transactions per second. (ii) to improve latency for high contention hotspot data. Our protocol commits 16x-18x more transactions compared to the open-source geo-distributed database CockroachDB. Meanwhile the increasing number of data leaks and data losses motivated me to build database systems focusing on data security. My other past and ongoing projects aim to build a foundation for the next generation of data management systems that preserve data integrity and privacy while being both fault-tolerant and efficient. The goal of my future research is to build efficient and practical database systems that meet increasing public expectations and legal requirements for data privacy that are currently lacking in existing systems.