K S Aravinda Kashyap

I'm pursuing my M.S. in Computer Science at Virginia Tech, where I work as a Graduate Research Assistant in the Security & Intelligence Lab, advised by Prof. Peng Gao.

My research sits at the intersection of AI security and safety and systems security, with a focus on the security of LLM-based agent pipelines as they move from research demonstrations to production deployments. I work on the failure modes that distinguish agentic systems from earlier LLM applications: indirect and cross-domain prompt injection, goal hijacking, MCP and plugin abuse, inter-agent trust escalation, session context contamination, and human-in-the-loop bypass. My current focus is on the attack and defense frameworks needed to evaluate such systems against realistic, reproducible adversaries.

Prior to graduate school, I was an Associate Software Engineer at Shell Plc and a Software Engineering Intern at the Indian Space Research Organisation (ISRO) on satellite telemetry systems for MEOSAR operations. I completed my B.E. in Information Science & Engineering from Siddaganga Institute of Technology, India.

I'd love to connect and chat about agentic systems, AI security, or systems engineering. Feel free to reach out!