Assistant Professor of Computer Science, Banaras Hindu University, Varanasi
Fulbright–Nehru Postdoctoral Fellow, University of California, Irvine · 2024–2026
I work on blockchain and privacy-preserving data systems — making decentralized ledgers efficient to query, and making the enormous volumes of sensor data around us usable without exposing the people they describe.
I am a Fulbright–Nehru Postdoctoral Research Fellow at the University of California, Irvine, hosted by Prof. Sharad Mehrotra in the Donald Bren School of Information & Computer Sciences. My work there sits at the intersection of blockchain systems and privacy-preserving data management: how to query decentralized ledgers efficiently, how to enforce access control inside them, and how to process sensor data from pervasive environments without compromising the privacy of the people it describes.
My home appointment is as an Assistant Professor in the Department of Computer Science (MMV) at Banaras Hindu University, Varanasi, where I have taught and led research since November 2020. I hold a Ph.D. and an M.Tech. in Computer Science & Engineering from Madan Mohan Malaviya University of Technology, Gorakhpur.
I am always glad to hear from students and collaborators working on decentralized systems, data privacy, or crowdsourcing.
How decentralized and data-intensive systems can be made efficient without giving up on privacy. Four threads run through the current work.
Query retrieval efficiency and rich queries over ledgers, access control inside the chain, and consensus mechanisms — including an empirical study of Ethereum’s transition to proof-of-stake under EIP-3675.
Inference-aware deletion in databases, and privacy-respecting query processing over heterogeneous sensor data in pervasive spaces.
Trustworthy task allocation and worker-set mapping for decomposable complex tasks, with results intended to fold back into working crowdsourcing platforms.
Compaction strategies for LSM-tree stores such as LevelDB, and pushing machine learning inside the database engine to accelerate analysis.
Current work at UC Irvine. With Prof. Sharad Mehrotra, I contribute to TIPPERS — a Testbed for IoT-based Privacy-Preserving PERvasive Spaces. The core problem is that heterogeneous sensors in a smart building generate enormous volumes of real-time data, and the difficulty is processing it efficiently at query time without surrendering the privacy of the people it describes.
Related threads include data processing in decentralized ledgers over geographically distributed data, and pushing machine learning inside the database engine to speed up analysis. I also mentor undergraduate researchers building a model that predicts Ethereum price movements by combining on-ledger data with external sources.
Recent work is listed below. The complete record, including earlier work on distributed real-time database systems, is on Google Scholar and DBLP.
Profiled on my arrival at UC Irvine, on joining Prof. Sharad Mehrotra’s group to work on the TIPPERS project, data processing in decentralized ledgers, and embedding machine learning inside databases.
“…envisioning productive research collaborations in the long term.” Sharad Mehrotra, Professor of Computer Science, UC Irvine
Asking me for a recommendation letter? Please read Before You Ask for a Letter first — it sets out who I can write for, how much notice I need, and what to send in your first email.