Headline

Founding Engineer @SecNode | Security Researcher | AI/ML Security | Whitebox Pentesting

About

Security researcher and founding engineer at SecNode focused on offensive security, AI/ML systems, and whitebox vulnerability research. My work revolves around understanding where complex systems fail especially across trust boundaries, multi-agent environments, parser logic, and modern AI infrastructure. I spend most of my time building research labs, developing deterministic PoCs, analyzing model behavior, and exploring how subtle architectural assumptions evolve into real-world security impact. Areas of interest: • AI/LLM Security • Application & Infrastructure Security • Trust-Boundary Failures • Mechanistic Interpretability • Multi-Agent Systems • Offensive Security Research

Location

Bengaluru

AI stack

claude

codex

cursor

hermes agent

hugging face

langchain

Work experience

Founding Engineer at SECNODE, Dec 2025 - Present · 9 mos, Founding Engineer focused on building autonomous offensive-security infrastructure powered by coordinated LLM agents. Leading development of multi-agent systems for: • penetration testing • attack surface analysis • vulnerability triage • threat intelligence workflows Architected orchestration pipelines using LangGraph, MCP integrations, and adaptive multi-model routing while researching AI-agent trust boundaries, coordination failures, and autonomous security reasoning reliability.

AI Security Researcher at SECNODE, Nov 2025 - Dec 2025 · 2 mos

Education

Bachelor of Engineering - BE, Computer and Information Sciences and Support Services, Dayananda Sagar College of Engineering, BANGALORE

Public links

github: https://github.com/av7danger

av7danger.dev: https://av7danger.dev/