Gemography
Full-Stack Engineer
January — July 2018
Backend systems and admin dashboards on Node.js, Vue.js and MongoDB, with geospatial proximity search, multi-channel authentication, and a production environment configured and maintained end to end.
- Geospatial
- Proximity search built on MongoDB
- 3 methods
- Email, phone and SSO authentication
- CI/CD
- Automated pipeline on a secured VPS
Backend systems and dashboards
I built backend systems and the administrative dashboards on top of them, using Node.js, Vue.js and MongoDB: scalable APIs paired with interfaces the team could actually operate.
Finding what is nearby
Proximity mattered to the product, so search was built on MongoDB’s geospatial features. That fed into customisable feeds ranked by both interest and proximity, rather than recency alone.
Getting people in the door
Authentication covered email, phone number and SSO, with phone numbers verified through Twilio. Each route had to land the user in the same place, reliably.
Documentation as a deliverable
The API was maintained as a fully documented Swagger surface, kept current rather than written once. The same standard applied to the codebase itself: clean and well documented, so the next person could move without archaeology.
Owning production
I configured and maintained the production environment on a DigitalOcean instance, and set up the CI/CD cycle using GitHub Actions and PM2 on a secured VPS, taking a merged pull request to a running process without manual steps in between.
Stack
- Node.js
- Vue.js
- MongoDB
- Swagger
- Twilio
- DigitalOcean
- GitHub Actions
- PM2