Building My First AI Cyber Guardian Agent: Reflections from the Google x Kaggle Intensive

Arul Mozhi — Sat, 13 Dec 2025 03:27:51 +0000

This is a submission for the Google AI Agents Writing Challenge: Learning Reflections

Hi, I’m Arulmozhi M, a first-year B.Tech CSE student specializing in Cyber Security at SRM University. I’m deeply interested in cybersecurity, AI, and building practical tools that make the digital world safer for everyone. Participating in the Google AI Agents Intensive with Kaggle has been one of the most transformative learning experiences of my academic journey. I joined the program with curiosity, but I walked away with clarity, confidence, and a fully working AI agent that I built from scratch.

🚀 Why I Joined

As someone passionate about cybersecurity, I’ve always wanted to build tools that help people stay safe online. When I saw this intensive, I knew it was the perfect opportunity to learn how AI agents actually work — not just in theory, but in real, hands‑on ways.

🧠 What I Learned

The biggest shift for me was understanding that AI agents are not just chatbots. They are structured systems with:

Main agents
Sub‑agents
Tools
Workflows
Policies and reasoning steps

Some concepts that really clicked for me:

✅ Modular agent design

✅ Tool‑calling and workflows

✅ Building responsibly with user trust

✅ Turning ideas into working prototypes

🛡️ What I Built: AI Cyber Guardian Agent

For my capstone project, I built an agent designed to educate, protect, and empower users in cyberspace — especially beginners who may not understand digital risks.

My agent includes:

Q&A Sub‑Agent – Answers common cyber safety questions
Password Strength Checker – Evaluates passwords with clear feedback
Password Generator – Creates strong, policy‑based passwords
Privacy Coach – Tips for Wi‑Fi, UPI, Aadhaar, and more
Quiz Engine – India‑specific cyber safety questions
Visualization – Brute-force resistance of passwords

🔧 Challenges I Faced

I struggled with:

Structuring multiple sub‑agents
Managing dependencies in Kaggle
Designing a clean workflow
Making responses beginner-friendly

But each challenge taught me something new — and the community, examples, and course materials helped me push through.

🌱 How This Changed Me

Before this intensive, AI felt like a huge, complicated world.

Now, it feels like a tool I can use, not something I’m intimidated by.

I learnt:

How to think like an AI engineer
How to design systems, not just code
How to build responsibly
How to turn ideas into working prototypes

Most importantly, I gained confidence that I can build real AI tools—even as a first‑year student.

💬 Final Thoughts

I’m grateful to Google, Kaggle, and the entire community for making this experience accessible, practical, and inspiring. This intensive didn’t just teach me AI—it opened a door to a future I’m excited to explore.

🔗 My Project Notebook

View my Kaggle notebook here

👤 About the Author

I’m Arulmozhi M, a first-year B.Tech CSE (Cyber Security) student from SRM University. I love exploring cybersecurity, AI agents, and real-world applications of technology. This was my first time building an AI agent from scratch—and it definitely won’t be the last.

🏷️ Tags

#googleaichallenge #ai #agents #devchallenge #cybersecurity #studentprojects #kaggle

DEV Community: Arul Mozhi