My research focuses on the development of trustworthy, dependable, and verifiable intelligent systems. I am particularly interested in combining machine learning and language models with formal methods, runtime assurance, and system-level reasoning for applications in cyber-physical and safety-critical environments.
Methods and architectures for making AI-assisted decisions more reliable, explainable, constrained, and suitable for use in systems where incorrect decisions may have significant consequences.
Application of formal verification, model checking, testing, and runtime assurance techniques to learning-enabled and cyber-physical systems, with particular emphasis on safety and dependability.
Integration of data-driven models with symbolic knowledge, constraints, verification mechanisms, and structured reasoning to build more dependable AI systems.
Architectures and decision mechanisms for privacy-conscious, stateful, and trustworthy use of language models and AI services in enterprise and distributed computing environments.
Machine learning, language models, predictive maintenance, embedded and edge AI, and intelligent decision support for engineering applications.
Keywords: Trustworthy AI · Safe AI · Formal Verification · Runtime Assurance · Cyber-Physical Systems · Neuro-Symbolic AI · Language Models · Dependable Systems · Privacy-Aware AI