Keyur Parag Joshi

Google Sunnyvale MP2,
1175 Borregas Ave,
Sunnyvale, CA 94089 USA

About Me

I am a Software Engineer (L4) at Google. My work focuses on evaluating the accuracy of AI agents when they perform multi-step tasks and use developer-defined functions and resources.

I recently completed my PhD at the School of Computing and Data Science at the University of Illinois Urbana‑Champaign. My advisor was Associate Professor Sasa Misailovic. I was previously a research intern at Microsoft Research India working with Rahul Sharma. I graduated from the Indian Institute of Technology Hyderabad in 2017.

Compositional Analysis of the Effects of Uncertainty on Computations
Keyur Joshi
PhD Dissertation
Full Text
GAS: Generating Fast and Accurate Surrogate Models for Autonomous Vehicle Systems
Keyur Joshi, Chiao Hsieh, Sayan Mitra, Sasa Misailovic
International Symposium on Software Reliability Engineering (ISSRE 2024)
Preprint
FastFlip: Compositional Error Injection Analysis
Keyur Joshi, Rahul Singh, Tommaso Bassetto, Sarita Adve, Darko Marinov, Sasa Misailovic
In Submission
Preprint

Verifying Controllers with Vision-based Perception Using Safe Approximate Abstractions
Chiao Hsieh, Yangge Li, Dawei Sun, Keyur Joshi, Sasa Misailovic, Sayan Mitra
Embedded Software (EMSOFT 2022)
Full Text with Appendix

Diamont: Dynamic Monitoring of Uncertainty for Distributed Asynchronous Programs
Vimuth Fernando, Keyur Joshi, Jacob Laurel, Sasa Misailovic
International Conference on Runtime Verification (RV 2021)
Full Text | Appendix | Conference Talk (YouTube)
ApproxTuner: A Compiler and Runtime System for Adaptive Approximations
Hashim Sharif, Maria Kotsifakou, Yifan Zhao, Akash Kothari, Ben Schreiber, Elizabeth Wang, Yasmin Sarita, Nathan Zhao, Keyur Joshi, Vikram Adve, Sasa Misailovic, Sarita Adve
ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (PPOPP 2021)
Full Text

Aloe: Verifying Reliability of Approximate Programs in the Presence of Recovery Mechanisms
Keyur Joshi, Vimuth Fernando, Sasa Misailovic
IEEE/ACM International Symposium on Code Generation and Optimization (CGO 2020)
Full Text | Appendix | Slides

Statistical Algorithmic Profiling for Randomized Approximate Programs
Keyur Joshi, Vimuth Fernando, Sasa Misailovic
ACM/IEEE International Conference on Software Engineering (ICSE 2019)
Full Text | Slides
Verifying Safety and Accuracy of Approximate Parallel Programs via Canonical Sequentialization
Vimuth Fernando, Keyur Joshi, Sasa Misailovic
ACM SIGPLAN Conference on Object-Oriented Programming, Systems, Languages and Applications (OOPSLA/SPLASH 2019)
Full Text | Appendix | Slides
ApproxHPVM: A Portable Compiler IR for Accuracy-Aware Optimizations
Hashim Sharif, Prakalp Srivastava, Muhammad Huzaifa, Maria Kotsifakou, Keyur Joshi, Yasmin Sarita, Nathan Zhao, Vikram S. Adve, Sasa Misailovic, Sarita Adve
ACM SIGPLAN Conference on Object-Oriented Programming, Systems, Languages and Applications (OOPSLA/SPLASH 2019)
Full Text
Identifying Optimal Parameters for Randomized Approximate Algorithms
Vimuth Fernando, Keyur Joshi, Darko Marinov, Sasa Misailovic
Workshop on Approximate Computing Across the Stack (WAX 2019) (Co-located with PLDI 2019)
Full Text | Slides