DB Lab
MongoDB, PostgreSQL and system design — from first principles to interview depth, in English and Hindi.
- Type
- Learning platform · bilingual
- Context
- Built independently
- Stack
- React · TypeScript · Vite · Tailwind CSS


01The problem
Database knowledge for interviews is scattered and shallow: definitions without the why, and examples that never run on a real database.
02The solution
A static learning app with 36 lessons and 314 interview questions. Every lesson follows one arc — story, why it exists, a real example, pitfalls, interview answers and a 10-second recap — and every example was verified on real MongoDB and PostgreSQL.
03Architecture
- Content as Markdown + JSON, validated at build time by a Vite plugin that fails with the file name
- One lazy chunk per lesson per language, so readers download only the lesson they open
- Bilingual UI and content: a type-checked Hindi dictionary and a translation checker that guards code and diagrams
- No backend: progress lives in localStorage, with export and import as a backup
04Hard parts
- Keeping 72 translated files faithful: code, diagrams and structure must stay identical to English.
- A fast first load with hundreds of pages of content — solved with build-time validation and per-lesson chunks.
05My contribution
your role on DB Lab
06Where it stands
36 lessons, 314 interview questions, two languages, deployable as a static site.
07Built with
Next project · 05
Business services described as data, not rebuilt as code.Service Studio →