Headline

Breaking AI apps so they dont break users | Web, API & AI Security | Contributor, Microsoft PyRIT & Prowler | Ex-Security Advisor @ GPTree

About

Most people see "hacker" and think of a guy in a hoodie. I think of it as asking one question: what happens if someone uses this the wrong way? That question took me from 50+ labs on TryHackMe and PortSwigger to testing a real AI product. The product was GPTree, an AI platform where you branch conversations and compare models. I tested the app and its APIs, wrote up what I found with ways to fix it, and sat down with the founder to discuss every observation. He added one of my ideas to the product roadmap. That taught me what no lab can: finding a bug is only half the job. Explaining it so someone actually fixes it is the other half. Here's what I do: → Web & API security testing (Burp Suite, Nmap, Nessus, Metasploit) → OWASP Top 10: SQLi, XSS, CSRF, broken access control → AI application security and LLM pentesting → Clear reports with fixes, aligned with ISO 27001 and NIST I'm a cybersecurity grad, and I share what I learn on Medium and GitHub, including the mistakes. Right now I'm looking for an Application Security role where I can test real products with a team that cares about security. If that's you, message me. Let's talk.

Location

Bengaluru

AI stack

aistudio

antigravity

claude code

claude

codex

cursor

Work experience

External Advisor, Security & Beta Feedback at GPTree, Sep 2025 - Sep 2026 · 1 yr 1 mo, Bengaluru, Most people see AI as a chatbot. I see it as a massive coordination problem. I’ve joined GPTree as an External Advisor because they are solving the one thing that kills AI productivity: Context Drift. I’m not here to just test a product. I’m here to help build a System of Record for AI logic. My focus is simple but critical: Securing the Logic: I’m red-teaming the Graft branching systems to ensure that as companies scale their AI agents, their data and logic remain bulletproof. Architecting the Journey: Working directly with the founder to refine how agents and humans interact in a multi-branch environment. If it’s not seamless, it’s not scalable. Expanding the Ecosystem: Leading the charge in India to connect founders and engineers with a tool that actually understands how complex thinking works. We aren't just launching a tool on June 1st; we’re changing how the world interacts with LLMs. If you’re still using linear chat, you’re playing the old game. We’re building the new one.

Offensive Security Training at TryHackMe, Mar 2025 - Sep 2026 · 1 yr 7 mos, India

Cyber Security Intern at Future Interns, Feb 2025 - Mar 2025 · 2 mos, Bengaluru, Karnataka, India, - Gaining hands-on experience in penetration testing, vulnerability assessment, and security analysis. - Working with Kali Linux, SIEM tools, and ethical hacking frameworks. - Performing security audits and responsible vulnerability disclosure. - Contributing to cybersecurity research and documentation.

Cyber Security Analyst at StableCount, Jul 2024 - Jan 2025 · 7 mos, Goa, India

Education

Bachelor of Technology - BTech, Computer Science, Presidency University Bangalore

Central Board of Secondary Education, 10th, Central Public School Mehkar

Public links

github: https://github.com/tejas0077

medium: https://medium.com/@tsaubhage0007