Aviral Shrivastava

Aviral Shrivastava is a professor at Arizona State University whose research focuses on compilers, machine learning systems, and hardware–software co-design for emerging computing platforms. His work spans ML accelerators, embedded and cyber-physical systems, and reliability-aware computing, with particular attention to making machine learning models efficient, robust, and deployable on constrained hardware. He has contributed to domain-specific languages and compiler infrastructures such as MLIR-based frameworks for digital signal processing, as well as techniques for protecting ML models from soft and hard errors, adversarial attacks, and out-of-distribution failures. His research also explores systems for connected autonomous vehicles, cooperative sensing, and responsibility-sensitive safety, bridging control, sensing, and learning in large-scale cyber-physical environments. More recently, he has investigated quantum and quantum-inspired machine learning models and data representations, highlighting future directions for specialized models and next-generation computing. His work appears in venues including ACM SIGPLAN/SIGBED conferences, IEEE Embedded Systems Letters, IEEE Access, ACM Transactions on Cyber-Physical Systems, and leading systems and architecture forums, reflecting a strong emphasis on practical, system-level AI.