It probably started with Siri
I grew up in India, and one of the first pieces of technology that genuinely confused me, in a good way, was my dad's first iPhone.
I was probably in seventh or eighth grade when I started messing around with Siri.
Today it feels completely normal. You ask your phone a question, tell it to call someone, set an alarm, whatever.
Back then, I thought it was kind of crazy.
You could just talk to this thing.
And somehow it understood what you were trying to say.
I remember wondering how that worked. How does a computer hear someone speaking, figure out what they mean, and then actually do something with it?
I obviously wasn't thirteen years old reading ML papers.
I had no clue.
I was just curious.
That was probably my first introduction to AI, even though I didn't really know what AI meant yet.
Later, computer science gave me a second set of questions.
What actually happens after I click a button on a website?
Where does all that data go?
How does YouTube serve the same video to millions of people?
How do systems that big not constantly fall apart?
That combination has pretty much stuck with me.
Part of me wants to understand how a machine can do something intelligent. The other part wants to know what has to exist around it to make the whole thing actually work.
India to the U.S.
I had wanted to experience life outside India for a long time.
When I was 16, I came to the United States.
I wouldn't describe it as some dramatic "I left everything behind and moved across the world" story. India is still home. My family is there. Getting on a plane didn't suddenly change that.
But coming here that young definitely changed me.
I was figuring out a different country, different people, a different way of living, and eventually college here too.
I ended up studying Computer Science at George Mason University.
A lot happened in those years.
Some of it was computer science.
A lot of it was just growing up.
I spent college saying yes to things
I never really had one perfectly planned lane.
Backend sounded interesting, so I tried backend.
Then data engineering. Machine learning. LLM research. Computer vision. RAG. Agents. Teaching. Hackathons.
If something made me curious enough, I usually found some excuse to get into it.
At Wi-Tronix, I worked with railway camera data and learned that "data engineering" feels very different when data is arriving every day and your code actually has to keep up with it.
At Miraivant, I worked on recruiting software with AI in the middle of it and realized that the impressive AI demo is usually only a small part of the actual product.
There's all this boring-looking stuff around it: APIs, databases, jobs, state, retries, failures, deployments. If those pieces don't work, the intelligent part doesn't really matter.
Then I got into LLM research and ended up much closer to inference itself than I expected.
I also spent several semesters teaching programming at Mason.
That one surprised me.
When someone is completely stuck, you can't just grab the keyboard and fix their code. You have to figure out what they think the program is doing.
Sometimes explaining a bug is harder than fixing it.
I got much better at both because of that.
And then there were hackathons.
I love hackathons. They're slightly ridiculous.
You show up with a few people, convince yourselves that an idea that should probably take two months can be built in a weekend, sleep very little, change the idea a few times, panic near the deadline, and somehow end up with something you can demo.
Sometimes it's genuinely good. Sometimes it barely works. Both are fun.
They taught me to stop thinking forever and just build the damn thing.
The part I'm still not satisfied with
I've built a lot of things.
Some for work. Some for research. Some for hackathons. Some because I had an idea at 1 AM and apparently decided sleep was less important.
And there are things I've taken further than what's on this site.
I've built prototypes privately, shown them to people who might actually use them, talked to potential users about the problem, and gotten far enough that I could imagine turning a few of them into something real.
And then...
I didn't.
Sometimes I lost confidence in the idea.
Sometimes I kept building instead of actually putting it in front of enough people.
Sometimes I convinced myself it needed one more thing before I could release it.
The repos are still sitting there.
That bothers me a little.
Because there's a difference between learning how to build software and learning how to build something people actually choose to use.
I've spent a lot of time learning the first part. Now I want to learn the second.
I want to put something into the world that strangers find without knowing me, use because it solves a real problem, and come back to because it's actually useful.
Not because it looks good on my GitHub.
Not because there's a hackathon deadline.
Not because I need another project on my resume.
Because somebody genuinely wants it.
That's probably the next chapter for me.
And ten years from now?
No clue.
Seriously.
I don't know where I'll be, what I'll be working on, or what technology will even look like by then.
But someday, I want to build my own company.
Not just so I can call myself a founder. I want to find something worth spending years on.
Build the product. Find the first people who care. Get things wrong. Fix them. Build a team. And see how far we can take it.
I'd love for it to become something genuinely big, probably much bigger than anything I can realistically see from where I'm standing today.
Maybe I get there on the first attempt.
Probably not.
For now, I'm trying to become a much better engineer, work around people I can learn from, and be ready when I eventually find a problem worth going all in on.
I don't want to aim for the highest thing I can already see.
I'd rather find out what's beyond it.