I am a second year college student. I have exams. I have assignments. I have a social life that I keep telling myself I will get back to someday.
And somewhere in the middle of all of that, I decided to build a startup.
I want to be honest about what that has actually looked like, because most founder content you read is either the polished success story or the dramatic burnout story. Mine is neither. It is messier than both.
The idea problem
The hardest part nobody warns you about is not the building. It is the not knowing whether what you are building is real.
I had the idea for Refactron sitting in my head for months before I wrote a single line of code. A safety-first refactoring tool. Something that would find problems in your code, fix them, and prove the fix was safe before touching the filesystem. I thought it was obvious. I thought people would immediately get it.
Then I started talking to developers about it and realised that most of them had just accepted that refactoring production code is risky and scary and something you do carefully or avoid entirely. They were not waiting for a tool. They had built their entire workflow around the absence of a reliable tool.
That was the moment I realised I was not solving a problem people were actively searching for. I was trying to create the realisation that the problem existed in the first place. That is a completely different thing and it is so much harder.
You spend weeks wondering if you are delusional. You write code at midnight and at 2am and you sit with your laptop in the hostel room between classes and you keep asking yourself some version of the same question: is this real or am I just doing this because I started and cannot stop?
I do not have a clean answer to that question even now.
What building alone actually feels like
I am the solo founder of Refactron. There is no co-founder to sanity check ideas with. No one to say "yeah that makes sense" or "that is a bad idea and here is why." Just me, a repository, and a growing list of things that need to get done.
Some days that feels like freedom. Most days it just feels like weight.
The codebase grew to 51 files, 135 tests, 96.8% test coverage. I built 8 rule analyzers, 5 complex refactorers, 14 auto-fixers, a backup and rollback system, a CI/CD gateway, a full CLI with 10 commands. I shipped repo connect. The library went live on PyPI with a great response of 3.5k+ downloads with 10–12 active users.
And after all of that I still had to build the verification engine, which is the entire point of the product.
There is something demoralising about finishing something and realising you are still at the beginning of what matters. You do all this work and the real work has not started yet. That is just how it goes and nobody tells you that either.
The applications
I applied to Y Combinator. I put real effort into it. I answered every question honestly. I talked about what I was building and why and what I had done so far.
Still got rejected which was as expected.
I applied to Lightspeed. I spent time thinking about what I was building at a level I had not before. What problem does Refactron solve. Who is the customer. Why am I the person to build it. What is the unique insight. Writing those answers down properly made me understand my own product better than I had before. That was the unexpected part. The application process was more useful than I expected regardless of the outcome.
Then something happened that I did not see coming.
India Ascends
I got accepted into India Ascends by Lightspeed.
I am going to be honest. When I saw the acceptance I sat with it for a moment not quite believing it. I had been working on this thing in college, mostly alone, mostly uncertain, and here was someone external saying: we think this is real.
India Ascends is Lightspeed's programme for early stage Indian founders. It is not just funding. It is access to a community of people who are doing what you are doing, who understand what it actually feels like, who have been through the questions that keep you up at night.
I do not have words for how much that matters when you have been mostly working alone.
The F6S thing
Around the same time I got some founder grant money through F6S.
This is the part of the story that feels almost surreal to write. F6S is a platform for startup founders and they surface grants and opportunities. I had signed up and filled out my profile and honestly kind of forgotten about it.
Then I got a notification. A small grant. Founders money.
It is not a huge amount. It is not the kind of thing that changes the trajectory of the company. But it is the first external money that came in specifically because someone looked at what I was building and thought it was worth supporting.
That hits differently than you expect it to. It is not really about the money. It is about the signal. Someone outside your own head looked at your thing and said: this is real enough to put something behind.
Where things actually are right now
The Python MVP is shipped. It is live on PyPI. Developers are installing it and running it on their codebases.
The verification engine is still being built. The three checks that actually make Refactron what it says it is. Syntax validation. Import integrity. Test suite gate. Before any file is touched. That is what v1.1.0 is about and it is not shipped yet.
I am also working on the TypeScript rewrite. A new npm package. Language-agnostic architecture. Blast radius analysis, which is something new that I am genuinely excited about. It tells you how much of your codebase would be affected if a fix went wrong, computed from your call graph before anything is applied. No refactoring tool does that right now.
The HN account karma situation is ongoing. My submission was blocked because my account does not have enough history. I am building it.
The website SSR was broken and is now partially fixed. The blog exists. The sitemap is live.
None of this is finished. All of it is moving.
What I want to say to anyone reading this
If you are a student building something and it feels impossible, I want you to know that feeling is accurate. It is hard. The validation problem is real. The alone problem is real. The not knowing problem is real.
But I also want you to know that the small wins are real too. The acceptance email is real. The grant notification is real. The first install from someone you do not know is real.
You do not need everything to be figured out to keep going. You just need enough signal that the thing you are building is not completely made up.
I have that signal now. It took a long time to get there and the work is nowhere near done.
But it is real. Refactron is real. And I am going to keep building it.
If you want to try it
TypeScript and JavaScript (v2.0 npm CLI):
npm install -g refactron
Python / PyPI:
pip install refactron
Run an analysis in your repo:
refactron analyze .
And if you are building something hard and want to talk to someone who is also in it, find me on X at @OSherikar. I am not hard to reach.
refactron.dev · docs.refactron.dev · github.com/Refactron-ai