WorkAboutWriting
Back to writing
ProjectsDocsMar 20265 min read

PaperLoop Problem Space Discovery

Defining the original problem and coming up with PaperLoop

Shanjit Thokchom profile photo
Shanjit ThokchomProduct Manager · Agentic AI

The Question Paper Problem

Last year, I met several people working in education. Some were administrators and support staff, but mostly teachers, a few of them close friends. I was in the middle of a six-month product management course at Masai School at the time, learning the fundamentals. It was during exam season that I started hearing the same complaint repeatedly: teachers were spending enormous amounts of time writing, formatting, and editing question papers, nights and weekends, just to meet deadlines.

A Firsthand Look

I got a firsthand look at this when I started helping one of my closer friends with her chemistry papers. She would draft them by hand, send them over to me, and I would type them out. The chemical formulas alone were brutal. Even after getting used to building them in MS Word, my speed only improved slightly. It was painstaking work.

And it didn't end at the first export either. There would be corrections, swapped sections, last-minute changes. She'd call and ask me to quickly swap a Class X paper with a Class XI version, and I'd have to dig up the Word document since Google Docs simply doesn't support advanced chemistry notation, make the edits, and send it back. A lot of effort for what felt like it should be a minor task.

The Complexity Beneath

By this point, I knew a solution was worth building. What I didn't yet understand was the complexity underneath it. During competition analysis, I discovered that the digital infrastructure required to support most existing solutions was simply out of reach for the majority of Indian schools. Even institutions with "International" or "New Horizon" in their names were largely operating on the same old-school setup.

Smart Classrooms

Existed but remained a minority

State-Led Digitization

Tools existed but effectiveness was questionable

Usage Gap

Whether tools were actually being used effectively was a separate problem entirely

The Time Tax

When I began formally validating the problem, I found an international study indicating that teachers spend 5 to 10 unpaid hours, and these are conservative figures, creating question papers alone. In India, that burden is compounded by what I came to call the tech deficit. Typing is difficult enough on its own; advanced formatting tools are complicated on top of that. Even for passionate, motivated teachers, the energy drain is significant and only compounds across a full school year.

The Financial Dimension

The infrastructure gap also has a financial dimension that goes beyond just buying devices. Licensing proper typesetting software, maintaining a reliable internet connection across classrooms, and training staff to use any of it costs money that most schools in India are simply not allocating toward exam preparation.

Platform Assumptions

Not every school can equip its teachers with the tools or the preparation time to adopt a platform-based workflow, and students cannot reasonably be expected to sit exams on a screen

Content Complexity

Exams are not just multiple choice. A chemistry paper has long-form derivations. A math paper has construction steps. A language paper has essays

Design Limitations

Platforms built around objective-type digital inputs were never designed for the full range of what a real exam looks like

The result is that the handwritten draft remains the most practical and universal format for Indian educators, and nothing in the current market is built to meet them there.