PeopleofSuperintelligence

Researcher

Sanjit A. Seshia

Sanjit Seshia is the Cadence Founders Chair Professor at UC Berkeley, where he develops formal methods for dependable AI and robotic systems. He earned a B.Tech. at IIT Bombay in 1998 and a Computer Science Ph.D. at Carnegie Mellon in 2005, where his doctoral work supplied reasoning methods for the UCLID verifier. His work includes neural network specifications, the Scenic language for generating training and test scenes, and the VerifAI toolkit for analyzing systems that use machine learning. With Edward Lee, he coauthored Introduction to Embedded Systems: A Cyber-Physical Systems Approach, and he brought formal verification into online teaching through CPSGrader. He became an IEEE Fellow in 2018 and an ACM Fellow in 2020, and shared the 2021 CAV Award for work on satisfiability modulo theories.

Works
551
Citations
17,059
h-index
67

Publication numbers from OpenAlex, as of 28 October 2026.

Timeline

  1. 1998–2005

    Computer science degrees

    He earns a B.Tech. in Computer Science and Engineering at the Indian Institute of Technology Bombay in 1998 and an M.S. in Computer Science at Carnegie Mellon University in 2000. In 2005, he completes a Computer Science Ph.D. at Carnegie Mellon, advised by Randal Bryant. His thesis develops automated methods for arithmetic reasoning that provide computational engines for the UCLID and TMV verification systems.

  2. 2007–2008

    Early career research awards

    In 2007, he receives a Presidential Early Career Award for Scientists and Engineers. He receives an Alfred P. Sloan Research Fellowship in 2008.

  3. 2014–2017

    CPSGrader and a textbook

    In 2014, he offers the edX course EECS149.1x on cyber-physical systems, using formal verification for automatic grading through CPSGrader. In 2017, MIT Press publishes the second edition of Introduction to Embedded Systems: A Cyber-Physical Systems Approach, which he coauthors with Edward A. Lee. The book covers modeling, design, and analysis of systems that combine computation, networking, and physical processes.

  4. 2018

    Neural network specifications

    In 2018, he coauthors Formal Specification for Deep Neural Networks. The work examines how to specify the behavior that neural networks must satisfy and discusses the challenges of applying formal methods to them.

  5. 2018–2020

    IEEE and ACM fellowships

    He becomes an IEEE Fellow in 2018 and an Association for Computing Machinery Fellow in 2020.

  6. 2019

    Scenic and VerifAI

    In 2019, he coauthors the PLDI paper on Scenic, a probabilistic programming language that generates training and test scenes for machine learning systems such as those used in autonomous cars and robots. He also coauthors the 2019 VerifAI toolkit paper; the software uses simulation guided by formal models and specifications to analyze systems with AI and machine learning components.

  7. 2021

    The CAV Award

    In 2021, he shares the Computer-Aided Verification Award for contributions to the foundations of satisfiability modulo theories. His doctoral work on the UCLID verifier and decision procedure helps establish this field, which supplies tools for verifying software and hardware.

  8. 2022

    Toward Verified Artificial Intelligence

    In 2022, he coauthors Toward Verified Artificial Intelligence with Dorsa Sadigh and S. Shankar Sastry in Communications of the ACM. The article sets out five challenges and fifteen principles for building AI systems with guarantees that they meet mathematically specified requirements.

  9. 2024

    Creating the NeuS conference

    In 2024, he helps create the International Conference on Neuro-Symbolic Systems, known as NeuS, and serves on its steering committee.

  10. As of October 2026

    Cadence Founders Chair Professor

    As of October 2026, he is the Cadence Founders Chair Professor in UC Berkeley's Department of Electrical Engineering and Computer Sciences and an affiliated faculty member of Berkeley Artificial Intelligence Research. His research develops formal methods for dependable and secure computing, with a focus on cyber-physical systems, computer security, machine learning, and robotics.

Career

  1. NowUniversity of California, BerkeleyCadence Founders Chair Professor
  2. NowUniversity of California, BerkeleyAffiliated faculty member, Berkeley Artificial Intelligence Research

Education

  1. 2005Carnegie Mellon UniversityPh.D. in Computer Science
  2. 2000Carnegie Mellon UniversityM.S. in Computer Science
  3. 1998IIT BombayB.Tech. in Computer Science and Engineering

From the biography.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Sanjit A. Seshia yet. Send us a free-licensed photo.

  1. Sanjit A. Seshia | EECS at UC Berkeley www2.eecs.berkeley.edu. Checked 28 October 2026.
  2. Sanjit A. Seshia | Carnegie Mellon University Computer Science Department csd.cs.cmu.edu. Checked 28 October 2026.
  3. Sanjit Seshia's Home people.eecs.berkeley.edu. Checked 28 October 2026.
  4. Lee and Seshia, Introduction to Embedded Systems ptolemy.berkeley.edu. Checked 28 October 2026.
  5. Sanjit Seshia's Publications formal specification for deep neural networks people.eecs.berkeley.edu. Checked 28 October 2026.
  6. [1809.09310] Scenic: A Language for Scenario Specification and Scene Generation arxiv.org. Checked 28 October 2026.
  7. [1902.04245] VERIFAI: A Toolkit for the Design and Analysis of Artificial Intelligence-Based Systems arxiv.org. Checked 28 October 2026.
  8. Sanjit Seshia wins Computer-Aided Verification Award - EECS at Berkeley eecs.berkeley.edu. Checked 28 October 2026.
  9. Sanjit Seshia's Publications toward verified artificial intelligence people.eecs.berkeley.edu. Checked 28 October 2026.