Building Safer AI: Alignment and Robust Cyber-Physical Systems
Why the next generation of AI must be predictable, aligned, and physically grounded
Dec 4, 2025
Kundan Kumar
Research Scientist | Safe Reinforcement Learning, Multi-Agent Systems & LLM Agents
Hi! I’m Kundan Kumar, a Ph.D. candidate in Computer Science at Iowa State University. My research focuses on creating intelligent and adaptable AI systems for next-generation cyber-physical infrastructure, integrating deep reinforcement learning (DRL), multi-agent systems, large language models (LLMs), and computer vision.
I develop safety-critical DRL frameworks that incorporate domain knowledge and uncertainty, enabling reliable decision-making in complex environments. Recent projects include exploring transfer learning and enhancing adversarial resilience across systems. I also create LLM-integrated simulation frameworks for autonomous systems, combining perception, trajectory planning, and natural language reasoning.
Outside of research, I share insights on Substack and YouTube. I enjoy cooking and ice skating 🛼 in my free time.
Other Research Interests
LLM Reasoning & Agents
Agentic workflows (LangChain/LangGraph), tool-use, memory, retrieval, planning & reflection loops for robust multi-step reasoning. Statistical ML
Uncertainty quantification, probabilistic modeling, and data-driven inference in dynamic environments. Autonomous Perception & Control
Vision-based perception (detection, segmentation, sensor fusion) integrated with learning-based control and trajectory planning for autonomous systems.
Agentic workflows (LangChain/LangGraph), tool-use, memory, retrieval, planning & reflection loops for robust multi-step reasoning. Statistical ML
Uncertainty quantification, probabilistic modeling, and data-driven inference in dynamic environments. Autonomous Perception & Control
Vision-based perception (detection, segmentation, sensor fusion) integrated with learning-based control and trajectory planning for autonomous systems.
Explore My Work
Talks
Publications
News Highlights
[Dec 2025]
Participated in an AI Strategy and AI Safety program hosted by
BlueDot Impact
, focusing on long-term AI risk, governance, and responsible deployment of advanced AI systems.
[Sep 2025]
[Mar 2025]
