Krishna C. Kalagarla
Papers
3
Total Citations
23
H-Index
3
About
Krishna C. Kalagarla is a researcher specializing in reinforcement learning, formal methods, and optimal control of stochastic systems. His work sits at the intellectually rich intersection of Markov decision processes (MDPs) and formal specification languages, tackling the fundamental challenge of designing intelligent agents that are both reward-maximizing and provably safe or correct by construction. Kalagarla's most notable contributions involve developing principled frameworks for synthesizing optimal policies for MDPs under temporal logic constraints — particularly Linear Temporal Logic (LTL) and Signal Temporal Logic (STL). His 2021 paper on model-free reinforcement learning under STL specifications (11 citations) is his most impactful work to date, introducing a practical RL algorithm that guarantees a desired probability of satisfying complex temporal specifications without requiring a system model. Complementing this, his work on discounted-reward MDPs under LTL specifications (8 citations) extended the field beyond finite-horizon tasks, addressing a significant gap in prior literature. Collectively accumulating over 20 citations across his published work, Kalagarla's research provides foundational tools for deploying autonomous systems in safety-critical environments — making his contributions particularly relevant to robotics, autonomous vehicles, and cyber-physical systems. His work bridges theoretical rigor with computational tractability, offering a compelling research trajectory for students interested in trustworthy AI.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3