Yuvraj Pradhan
Independent Security Researcher
Independent Researcher
Yuvraj Pradhan is an independent security researcher and AI systems engineer focused on agentic AI security, workflow authorization, and trust-boundary failures in production systems. His current research investigates approval-provenance failures across 43 agentic AI, durable-workflow, CI/CD, and identity-governance systems, including coordinated disclosures to Microsoft, OpenAI, AWS, and other vendors. He is the first author of Springer Nature research on a 125M-parameter NanoLLM and has spoken at Linux Foundation MCP Dev Summit events in Bengaluru and Mumbai. His work combines offensive experimentation, systems-security research, and practical defenses for emerging AI infrastructure.