papers

context

in my school, every time an exam finished, one copy of the question paper was sent to the library archives for records.

the papers were there, but the workflow had annoying friction. if you wanted to practice past papers before periodic tests or pre-boards, you had to walk to the library, request the physical copy, and wait around at the copier machine to pay for photocopies out of pocket. once school hours ended, you had zero access.

i got sick of the friction. access to old question papers shouldn’t cost money or mean waiting around at a photocopy machine. so one day i decided to take the whole thing online and built the platform from scratch.

the pitch and rollout

i built the frontend, the edge storage layers, the faculty auth, and the in-browser reader. once the base version was running, i sent the link to our principal in a late-night linkedin dm.

she saw it a couple days later and her response was simple: let’s implement this.

i took the project to my computer teacher to figure out next steps, got the green light to coordinate with the administration, and that kicked off the rollout. from there, i wired up the faculty portal, added server-side hmac authentication so teachers could handle uploads without dealing with cloud consoles, and wrote the faculty guide.

right now, the platform is in its implementation phase, handed off to the computer department teachers to manage paper uploads.

i also kept the whole repository open source under the mit license so any other school can fork it and set up their own archive without paying outside vendors.

how it works

i built this around two practical constraints: zero friction for students, and zero cost for the school.

• zero friction for students: no signups, no logins, no paywalls, and no trackers. open the link on a phone, pick your class, and start reading in under three seconds. includes an in-browser reader with a one-click toggle between the question paper and the marking scheme.
• zero-rupee architecture: running on serverless edge compute with a three-tier fallback. it streams through a vercel blob proxy, falls back to cloudflare r2 for zero-egress object storage, and caches locally with indexeddb. the school pays literally zero rupees to host or run it.
• faculty portal: protected by hmac sha-256 session auth. teachers can drag and drop scanned papers, attach marking schemes, and manage terms without touching a database or cloud dashboard.
• academic invariants: hardcoded validation rules matching our school curriculum (classes 9th to 12th, science and commerce streams, periodic tests, and pre-boards) so papers never end up mislabeled.

what i learned

• navigating school bureaucracy: writing code was the fast part. pitching the idea, coordinating between the principal and the computer department, and handing off clean tools to non-technical teachers was the real lesson.
• building under zero-rupee constraints: making a production system survive on free-tier limits forced clean engineering: edge reverse-proxies for private blob assets, client-side vector pdf generation, and solid offline fallbacks.
• friction kills tools: if students had to create accounts or log in just to view a 10th-grade science paper, they would just go back to asking friends on whatsapp. removing all gates made it actually useful.