Interviews

The interview wasn't about my code. It was about my engineering journey.

An interview with Galaxy.ai, after an assignment where I hunted bugs in their product. I had real work to talk about. What I didn't do was turn that work into clear technical explanations, and I noticed the gap while I was still talking.

Bammidi Govinda RaoAuthor

Oct 9, 2026 · 6 min read

Follow on Medium
arc: a bug hunt assignment i liked → an interview that asked about my journey, not my code → real experience, explained too shallowly → no next round and no rejection yet → go back and learn my own systems properly

the assignment came first

this is my third interview entry. the first was a screening round where i had nothing to give. the second was a java full stack round i crammed one night for. this one was with galaxy.ai, for a backend focused engineering role.

before the interview there was an assignment. explore magica ai and find bugs.

i liked that. it wasn't a problem statement to solve. it was a real product, and the job was to use it, watch how it behaved, and notice where it didn't do what it was supposed to.

that is what got me interested in the role.

but the interview itself was different from what i had prepared for.

i had a lot to talk about

i introduced myself as a 2025 b.tech graduate and one of the founding engineers at māhita technologies, an early stage startup building software for schools.

we are a team of three, and i'm responsible for the core technical implementation. frontend, backend, database, deployments, testing, and product discussions. our founders are in canada, so i have also visited schools, spoken with principals, understood what they needed, and helped with sales.

our main product is a multi-tenant school erp, with modules for fees, marks, student attendance and other school operations.

we also built vinodam, a quiz platform that started from a real request from one of our schools. around 300 students took part in the event, and it worked.

none of this was built to fill a resume. it came from real requirements, real users, and having to make something work.

i was happy to talk about all of it.

having experience and communicating that experience are two different things.

the education question

he asked about my education, and whether i had taken jee.

i told him the truth. i got into engineering through a management seat, because i hadn't prepared well for the state entrance exam.

in college i got interested in software, started learning on my own, built projects, and eventually joined a startup. i graduated with an 8.17 cgpa and the jntugv gold medal for my branch.

i don't want to rewrite my past to make it sound perfect. i didn't have a perfect start. i kept learning, and i ended up responsible for software that real schools used.

that part matters more to me than pretending i always knew what i wanted.

the question i could have answered better

then he asked about our tech stack, our architecture, and our engineering team.

i said react, node.js and postgresql. i talked about my responsibilities, and how i had started focusing more on backend architecture and system design.

but i could have gone much deeper.

i could have explained how our multi-tenant architecture works. how we keep each school's data separate. how a request gets authenticated, and then authorized against the school it is asking for. why we made the backend decisions we made.

i have worked on all of that. i have debugged it, changed it, and been the one responsible for getting it into production.

i just didn't turn it into clear technical explanations in that conversation.

the strange part is that in my last interview, that exact architecture was the answer that came out without any effort, because the question asked for it directly. this time the question was broader, and i stayed at the surface.

Ex: "i worked on the backend" tells the interviewer that i was there. "this is why the backend is designed this way, this is what i considered instead, and this is what can go wrong" tells them how i think.

the second answer is the one that shows an engineer. that is what i need to get better at.

what vinodam taught me that tutorials couldn't

i talked about vinodam.

a principal wanted a quiz experience that would get students engaged. we turned that request into a separate product, built the quiz platform, and used firebase for the realtime part. around 300 students took part.

what it taught me is that engineering doesn't end when the code works on my laptop. someone has to understand the requirement, make the decisions, deploy it, handle what goes wrong, and make sure it works for the people actually using it.

working at a small startup gave me that responsibility early.

but owning a product doesn't mean i know everything about the systems underneath it. there is still a lot for me to learn about backend internals, how the database actually behaves, reliability, performance and system design.

i want to be able to explain not just what i built, but why it works the way it does.

why i wanted this role

i said in the interview that ai is making it easier to write software, but debugging and understanding systems still matter.

i use ai tools in my work. they help me move faster, try ideas, and build features.

but i don't want my engineering ability to depend on how good a model is at generating code. i want to understand the system well enough to spot a bad implementation, dig into behaviour that doesn't make sense, and make a call without blindly accepting whatever the ai gave me.

that is also why the assignment pulled me in. it was about looking hard at a real product, not writing code for a problem statement.

and it's why i'm getting more interested in backend architecture and system design. i want to understand the systems i build at a deeper level.

how it ended

he asked my notice period. i said two weeks.

he said he would go through my resume in more detail and see whether my experience matched what they needed. if it did, they would share next steps.

i asked what the next round would be. a technical round, he said.

that was the end.

so right now there is no confirmed next round, and no rejection either. i'm waiting.

i want to write that down honestly, instead of turning every interview into a success story or a failure story before i actually know what happened.

what i'm taking from it

i have worked on real products. i have dealt with customers, built features, deployed them, and taken on responsibility beyond writing code. i should be confident about that.

but confidence isn't enough. i need to explain my technical decisions clearly, and be ready for the details behind the systems i have built.

so the next step is simple. go back to my own projects. draw their architectures. understand the trade-offs i made. practise explaining them without falling back on vague descriptions.

i don't want to learn system design just to give better interview answers. i want to understand my own products better.

whether or not this turns into a next round, it gave me something concrete to work on.

i have spent time learning how to build software. now i need to get better at explaining the engineering behind it.

#buildinpublic #softwareengineering #systemdesign #backend #interviews #learninginpublic

Send this as proof →Share on LinkedIn