How I Actually Got Here

Personal blog

It all started around 2018, if not earlier, when I stumbled upon a C++ tutorial by CodeWithHarry on YouTube. I still don't know exactly what sparked my curiosity, but something clicked. The next challenge was obvious. How was I supposed to practice without a computer?

At the time, I didn't own a PC, laptop, or even a personal phone. My siblings and I shared a 2GB RAM Android phone. After a bit of research, I discovered Termux. I watched tutorials on how to install a C++ compiler inside it and started solving basic programming problems directly on my phone. My excitement only lasted a couple of days before fading, but that brief experience planted the seed of my interest in technology.

Then came 2020.

COVID hit, lockdowns started, and suddenly I had something I had never had before, unlimited free time and a personal phone for online classes. Jackpot.

I spent more and more time on Discord. At first, it was just hanging out with friends, but eventually I discovered NFTs, Web3, and the metaverse. It felt like an entirely new world.

I minted a few free NFTs and eventually started doing marketing for NFT founders. Surprisingly, it earned me quite a bit of money, at least for a 15-year-old. While working with these communities, I started building Discord bots, web scrapers, and webhook-based monitoring systems for founders. Everything I learned came from Stack Overflow, Discord communities, YouTube, and online forums. Along the way, I picked up Python and learned the basics of MongoDB.

That was essentially my NFT phase. The NFTs eventually faded away, but my curiosity for building things didn't.

Then ChatGPT was released.

Initially, I used it for school assignments and random questions like everyone else. But one day, while building a Discord bot, I got stuck on an error that I simply couldn't solve. I uploaded a screenshot of the error to ChatGPT.

It generated the corrected code.

I copied it, ran it, and it worked immediately.

I was genuinely blown away. I couldn't believe an AI could debug my code like that. I proudly showed the bot to my friends, and for me, it felt like magic. Eventually I ran into ChatGPT's early limitations. It wasn't nearly as capable as today's models, but that experience completely changed how I looked at AI.

During 11th and 12th grade, I still had a lot of free time because I rarely attended school. Looking back, I definitely wasted a fair bit of it, but I also learned a lot.

I taught myself HTML, CSS, and JavaScript and built a handful of simple websites. I learned Git and GitHub and became comfortable using version control.

By 12th grade, I had developed a strong interest in cybersecurity. I started exploring virtual machines, Linux, and different distributions like Kali, Ubuntu, and Arch. I experimented with penetration testing tools, learned some Bash scripting, and spent countless hours exploring how systems worked under the hood.

That interest is actually why I chose Cybersecurity as my specialization in college. Even today, it is something I am still passionate about.

I somehow survived 12th grade, barely. Mathematics absolutely destroyed me. After that, I enrolled in a Computer Science degree.

My first semester was mostly about adjusting to college life. I built tiny projects, a calculator in C, a to do app, and other beginner projects that I would rather pretend never existed.

During my second semester, I learned the basics of Data Structures and Algorithms and improved my Python skills, but overall I felt like I had wasted my entire first year of college.

Then came the semester break.

I decided something had to change.

While browsing YouTube, I came across Blackbox AI. I installed it inside VS Code, built a portfolio website in a single day, and from that point onward there was no looking back.

I started exploring every AI coding assistant I could find, Trae, Windsurf, Qodo, Warp, Cursor, and many others. I loved how much faster I could build software. Instead of spending hours searching documentation, I focused on understanding problems and building solutions.

Throughout my third semester, I became obsessed with learning how to work effectively with AI. I learned prompt engineering, experimented with modern AI tools, and also explored MLOps concepts like Docker, Kubernetes, Jenkins, CI/CD pipelines, and a bit of AWS.

After my third semester ended, another turning point came.

A friend introduced me to AI automation.

I discovered n8n and started building real world automation systems using tools like Supabase, Retrieval Augmented Generation (RAG), vector databases, APIs, and LLMs. Before long, I had moved beyond simple automations into Agentic AI.

I started exploring Codex, Claude Code, and many other AI agents while experimenting with different frameworks, orchestration systems, and developer tools. I built AI agents, lead generation systems, GTM pipelines, workflow automations, and business focused AI applications.

That brings me to where I am today.

Looking back, my journey has never followed a traditional path. I never learned technology through structured courses or classrooms. Most of what I know came from curiosity, experimenting, breaking things, fixing them, and building projects that solved real problems.

Every phase, whether it was coding on a shared Android phone through Termux, building Discord bots during the NFT boom, exploring cybersecurity, or now developing Agentic AI systems, has been driven by the same thing.

Curiosity.

And I do not think that is going away anytime soon.