I am a Machine Learning Engineer & Researcher specialized in building responsible, human-centered AI. I bridge the gap between complex research (MARL, AI Safety) and production-grade engineering (Agentic RAG, Scalable MLOps).
- 🛡️ StepShield: Developing a step-level safety benchmark for autonomous LLM code agents to detect rogue behavior (exfiltration, resource abuse) in real-time. Targeting ICML 2026 publication.
- 🤖 Multi-Agent Systems: Surveying MARL applications in cybersecurity, focusing on attacker-defender dynamics and interpretability.
- 🎵 Generative Audio: Fine-tuning PyTorch-based models to generate personalized meditation/wellness audio.
- 🏗️ Agentic RAG: Architecting high-accuracy mental health chatbots using CrewAI and LangGraph with 95% safety safeguards.
Languages & Core: 
AI & MLOps Frameworks: 
Scale & Performance: 
