November 2020. A Singapore edtech startup I co-founded with Benjamin.
the problem
Singapore university students are exam-obsessed — but quality guidance is gated. Past year papers are publicly available; clear, worked explanations of the answers are not. You either knew someone who aced the module, or you winged it.
what we built
BellCurveHero was a two-sided marketplace:
Supply side — we recruited A+ students and teaching assistants and paid them to record video walkthroughs of past year exam papers. Verified experts, real explanations, module-specific.
Demand side — students at NUS and SUSS paid for access to the video library, browseable by university, module, and exam paper. We focused on STEM and business modules first.
The web app was built on Django with django-allauth for user authentication, Vimeo for video hosting, and Tailwind CSS for the frontend. Users could sign up, verify their email, and access the video library.
what happened
We got real users. Students were willing to pay.
I sank a few thousand dollars of my own money into the startup — production costs, hosting, recruiting content creators. The unit economics were starting to make sense.
The harder problem was the co-founder. Benjamin didn't have a clear grasp of the business model and wasn't pulling equal weight. After trying to work through it, I made the call to walk away rather than keep funding a partnership that wasn't working.
stack
Python, Django, django-allauth — web app and user auth. Vimeo — video hosting. Tailwind CSS — frontend styling.
Source on GitHub.