Innovation Hall E458
82 University Place
Burlington, VT 05405
United States
- Ph.D., Massachusetts Institute of Technology
- M.S., Massachusetts Institute of Technology
- B.S., Indiana University
Department of Computer Science
JOURNALISTS: Need help scheduling an interview? Contact UVM media relations or search for a UVM expert.
Areas of expertise
Security, privacy, applied cryptography, programming languages and formal verification, AI and machine learning; differential privacy, secure multiparty computation, zero-knowledge proofs, type systems, privacy-preserving ML
BIO
Joseph Near is an Associate Professor of Computer Science at the University of Vermont, where he is a member of the PLAID research group and the UVM Center for Computer Security and Privacy, and an affiliate of the Vermont Complex Systems Institute. His research sits at the intersection of security, privacy, and programming languages, with the goal of making it practical to learn from sensitive data without exposing it. He builds languages, analysis tools, and systems that make strong guarantees such as differential privacy, secure multiparty computation (MPC), and zero-knowledge proofs easier to use correctly, drawing on type systems, formal verification, applied cryptography, and machine learning.
His work includes Duet, a programming language that statically enforces differential privacy (ACM SIGPLAN Distinguished Paper, OOPSLA 2019), along with tools for privacy-preserving deep learning, secure aggregation, language-based security for MPC, and choreographic programming. He holds an NSF CAREER award on distributed differential privacy via secure multiparty computation, and his research has been funded by NSF, DARPA (including the Brandeis, SIEVE, and TOPSPIN programs), DOE, and Amazon. Recent projects also study the usability of differential privacy tools, privacy for large language models, and fairness in machine learning.
Near works closely with NIST on privacy engineering. He is a co-author of NIST SP 800-226, Guidelines for Evaluating Differential Privacy Guarantees, and has written and coordinated NIST’s blog series on differential privacy and privacy-preserving federated learning. He has also worked with Galois on auditing differential privacy mechanisms for the U.S. Census Bureau. He is the co-author of the open textbooks Programming Differential Privacy (with Chiké Abuah) and Differential Privacy for Databases (with Xi He), and the author of Programming MPC. He co-chaired the Theory and Practice of Differential Privacy (TPDP) workshop in 2024 and 2025 and serves on the program committees of venues including CCS, USENIX Security, PETS, and VLDB.
Before joining UVM in 2018, Near was a postdoctoral researcher at UC Berkeley with Dawn Song, where he worked with Uber’s privacy engineering team on open-source tools for differentially private SQL analytics. He received his Ph.D. and M.S. in Computer Science from MIT, advised by Daniel Jackson, where he developed static analysis and verification tools for finding security bugs in web applications and co-developed the Alloy* constraint solver (ACM SIGSOFT Distinguished Paper, ICSE 2015). He received his B.S. from Indiana University, where he worked with Daniel Friedman on logic programming. At UVM he teaches courses on data privacy, secure distributed computation, and compiler construction.
Courses
- CS 3110 — Data Privacy
- CS 3020 — Compiler Construction
Bio
Joseph Near is an Associate Professor of Computer Science at the University of Vermont, where he is a member of the PLAID research group and the UVM Center for Computer Security and Privacy, and an affiliate of the Vermont Complex Systems Institute. His research sits at the intersection of security, privacy, and programming languages, with the goal of making it practical to learn from sensitive data without exposing it. He builds languages, analysis tools, and systems that make strong guarantees such as differential privacy, secure multiparty computation (MPC), and zero-knowledge proofs easier to use correctly, drawing on type systems, formal verification, applied cryptography, and machine learning.
His work includes Duet, a programming language that statically enforces differential privacy (ACM SIGPLAN Distinguished Paper, OOPSLA 2019), along with tools for privacy-preserving deep learning, secure aggregation, language-based security for MPC, and choreographic programming. He holds an NSF CAREER award on distributed differential privacy via secure multiparty computation, and his research has been funded by NSF, DARPA (including the Brandeis, SIEVE, and TOPSPIN programs), DOE, and Amazon. Recent projects also study the usability of differential privacy tools, privacy for large language models, and fairness in machine learning.
Near works closely with NIST on privacy engineering. He is a co-author of NIST SP 800-226, Guidelines for Evaluating Differential Privacy Guarantees, and has written and coordinated NIST’s blog series on differential privacy and privacy-preserving federated learning. He has also worked with Galois on auditing differential privacy mechanisms for the U.S. Census Bureau. He is the co-author of the open textbooks Programming Differential Privacy (with Chiké Abuah) and Differential Privacy for Databases (with Xi He), and the author of Programming MPC. He co-chaired the Theory and Practice of Differential Privacy (TPDP) workshop in 2024 and 2025 and serves on the program committees of venues including CCS, USENIX Security, PETS, and VLDB.
Before joining UVM in 2018, Near was a postdoctoral researcher at UC Berkeley with Dawn Song, where he worked with Uber’s privacy engineering team on open-source tools for differentially private SQL analytics. He received his Ph.D. and M.S. in Computer Science from MIT, advised by Daniel Jackson, where he developed static analysis and verification tools for finding security bugs in web applications and co-developed the Alloy* constraint solver (ACM SIGSOFT Distinguished Paper, ICSE 2015). He received his B.S. from Indiana University, where he worked with Daniel Friedman on logic programming. At UVM he teaches courses on data privacy, secure distributed computation, and compiler construction.
Courses
- CS 3110 — Data Privacy
- CS 3020 — Compiler Construction