Building for 6,000 Schools: The SCERT Textbook Distribution Platform

By the time this project came around, I had already built the frontend for GSP. So when the next call came, it came from someone who had watched me deliver.

A few months after the GSP UI work, Diptanu — Aaharan's CTO — told me they had landed a new contract, and it was a big one: streamline the entire textbook distribution for roughly 6,000 schools across Tripura, for the SCERT. Imagine every school in a state raising requests for books, districts approving them, stock being drawn down, and physical shipments going out with paperwork — and imagine all of that running on paper and spreadsheets. That was the problem. Build the system that replaces it.

I said yes. And then I had to be honest with myself about something.

The project that made me partner up

Back then, I did not really know backend. I could build a beautiful, functional frontend all day, but a government supply-chain system for 6,000 schools needs a serious backend, and I was not going to fake my way through one at that scale.

So instead of pretending, I brought in a friend — Arijit. We split the work down the natural seam: I took the UI and the entire frontend, he took the backend, and we agreed on a revenue split up front. It was my first real collaboration on something this size, and it taught me a whole category of lessons that had nothing to do with code.

Because working with another person is its own engineering problem. You have to agree on the shape of the API before either of you builds against it. You have to keep two halves moving without blocking each other. And yes — there were disagreements along the way, the ordinary friction of two people with their own standards trying to ship one thing. We worked through them, and the project got done. I came out of it a better collaborator, not just a better developer. Knowing when to bring someone in instead of stretching past your depth is a skill I did not have before SCERT.

The workflow at the heart of it

The core of the system is a supply chain modeled in software:

• Requisitions — schools raise requests for the textbooks they need, inside a controlled requisition window so the whole state submits on the same clock.
• Review & approval — requests move up the hierarchy, gated by role: district-level and state-level officials see and act on what belongs to them.
• Stock — inventory is tracked against those requisitions so the system knows what can actually be fulfilled.
• E-challans — dispatch happens with electronic challans, the formal paperwork that says these books went to that school, generated as clean PDFs.
• Issue tracking — the whole flow is auditable from request to delivery.

Modeling that cleanly — role-based access, a requisition window that opens and closes, stock that has to stay honest against thousands of simultaneous requests — was the real design work. It is easy to build for one school. Building for six thousand, where a small mistake multiplies across a state, is a different discipline.

The stack reflected that seriousness: a modern Hono + Prisma backend with better-auth, a Vite + React frontend, and PDF generation for the challans and reports that officials actually need on paper.

Built to survive an audit

A government system for 6,000 schools does not just get deployed — it gets audited. Someone whose entire job is to break your security will go through it before it is allowed near real infrastructure.

So the later stretch of this project was about hardening it to that standard: reviewing the codebase and dependencies specifically for security gaps and closing them before launch, treating the audit not as a hurdle but as the bar the work had to clear. I am not going to detail the findings here — publishing the soft spots of a government platform would be exactly the wrong thing to do — but the short version is that we put the system through a rigorous security pass so it could stand up to the official one.

Where it stands

The platform is approved and currently under government security audit. It is done, it works, and it goes live to real schools the moment the audit is passed. Honest status: finished and waiting on the last, most important gate.

What I took from it

SCERT is the project where I stopped being only a solo builder.

It taught me to scope honestly — to admit I did not have the backend depth for something this large and to bring in the right person instead of bluffing. It taught me how to divide work, agree on contracts between two halves of a system, and push through the friction that comes with any real partnership. And it taught me what it feels like to build something where the stakes are a whole state's schoolchildren getting their books on time.

You do not learn that building alone. Sometimes the most important thing a project teaches you is who to build it with.