Pritam Sen PhD ’25 Authors Paper with Professor Cristian Borcea at ACM CCS Conference, Now Researches with Meta

Pritam Sen earned his Ph.D. in Computer Science at NJIT in May 2025, and has achieved what in horse racing is called the triple crown.

He most recently had a first-author paper at the highly competitive ACM CCS conference with his Ph.D. advisor, Professor Cristian Borcea. The paper presented CryptGNN, a security solution for third-party graph neural network models in the cloud. Prior to this, he was part of a research team led by Borcea that earned a U.S. patent in collaboration with AT&T. Now, his dream job as a research scientist with Meta has him working on recommendation models for designing Facebook Reels. 

CryptGNN is a machine learning as a service (MLaaS) platform where clients, such as small companies and startups, can use AI models trained on graph data by large companies, e.g., fintech, healthcare, social media, while maintaining the privacy of their data used for AI inference. 

Through the use of designed protocols for secure multi-party computation done in the cloud and then distributed to multiple servers from different providers, such as Google, Amazon, etc., large company model owners and clients can upload and share data without sensitive information leakage. 

The research team demonstrates that the solution works successfully, as it provides accurate results, is efficient, and preserves privacy. In terms of security and overhead, CryptGNN was tested using AWS servers and open-source datasets and ultimately determined to be provably secure. 

Sen and Borcea’s findings were recently presented at the ACM CCS 2025 conference in Taipei, Taiwan, which is the flagship annual conference on cybersecurity. The event brings together information security researchers, practitioners, developers, and users from all over the world to explore cutting-edge ideas and results.

Sen said, “When I was exploring schools for my Ph.D., I knew I wanted to be close to New York City. What ultimately brought me to NJIT was Dr. Borcea’s work with mobile and cloud computing. His expertise and experience were a perfect match for my interests and professional goals. I knew I would grow through his guidance and mentorship.” 

Borcea was equally effusive in his praise of Sen.

“Pritam is the type of Ph.D. student every professor wants to have: intelligent, creative, independent, and reliable. During his Ph.D. studies at NJIT, he made substantial contributions to the field of security and privacy for machine learning/AI. I’m very happy for his continued success, as demonstrated by our recent paper with ACM CSS, our U.S. patent done in collaboration with AT&T Research, and his job as Research Scientist with Meta,” he said. 

In speaking of his current position with Meta, Sen stated, “My work on designing recommendation models for Reels makes for a scalable system that is personalized and private. I couldn’t ask for a better job, and I extend my thanks to everyone, particularly Dr. Borcea, for giving me the opportunity to present my work at several conferences, which also helped my presentation skills and ability to write better. I think NJIT worked out very well for me!”