Conducting doctoral research on runtime governance and security mechanisms for autonomous AI agent systems. Focusing on Agentic AI, Multi-Agent Systems, adaptive control, and trustworthy execution of long-horizon agent workflows.

Mentoring undergraduate and graduate researchers within the SafeAgent research program. Providing guidance on research methodology, system design, implementation, experimental evaluation, and scientific writing.

Conducted research on runtime governance and security mechanisms for LLM-based agent systems. Developed SafeAgent, a runtime protection architecture for agentic systems, focusing on risk-aware decision making, agent safety, and defense against prompt injection, tool-use vulnerabilities, and workflow-level attacks.

Worked on AIGC systems and speech generation pipelines, including multi-speaker emotional TTS, model fine-tuning, text-audio alignment, and automated inference tools.

Investigated machine learning approaches for SQL injection detection using query modeling, feature extraction, and classifier evaluation. Developed research results presented at a student scientific conference.