Case study · Oct 2023 to Jun 2024
Baideshik Rojgar
The app Nepali migrant workers use to find jobs abroad. I led its mobile development to over a million installs.
1,000,000+ installs on Google Play at a 4.4 rating, for an audience on cheap Android phones and slow connections, with payments, ads and live radio added without breaking what worked.
The problem
Hundreds of thousands of Nepalis leave each year to work in the Gulf, Malaysia and beyond. The information they need, which jobs are real, what the process is, what their rights are, was scattered and often wrong. Baideshik Rojgar (also known as Sajha Jobs) is where they look.
The audience is the hardest kind to build for: entry-level Android phones, patchy mobile data, low tolerance for anything confusing, and a cost of error measured in someone's savings.
What I did
- Led mobile development of the Flutter app through its growth past a million installs, keeping the rating at 4.4.
- Added online payments through eSewa and Khalti, the two rails Nepali users actually have.
- Added ads and a live radio stream without hurting the core job-search flows.
- Set up automated releases so a fix reached users in hours rather than days.
What I learned
Scale is a product constraint, not a vanity number. At a million installs every rough edge is a thousand support messages, and every release is a risk. The discipline I use now, staged rollouts, automated release checks, a verification script before every upload, started here.
What I would do differently
- Add crash and performance monitoring earlier. We learned about slow screens from reviews before we learned from data.
Have something like this in mind?
Describe it to the fit assistant. It cites this case study where it applies and says where it does not.