Short answer: Next.js (or Remix) + Postgres + a managed deployment (Vercel, Fly, Render) + Stripe (via partner) — boring stack, fast time-to-first-customer.
Choices that hold for the first 18 months.
- Frontend: Next.js + TypeScript + Tailwind.
- Backend: Same Next.js API routes for small apps; tRPC or a Python service for heavier needs.
- Database: Postgres (Neon, Supabase, or Railway).
- Auth: Clerk or Auth.js.
- Payments: Stripe via partner; Khalti for Nepali users.
- Observability: Sentry + a logger.
The exception most miss. Choice of stack is rarely the failure mode. Distribution is. Spend 60% of week-one on the first ten customers, not on the framework debate.
Which technical choice would you reverse after shipping?
Discussion (0)
Sign in to comment, vote, and join the community.
Sign in Create accountNo comments yet. Start the discussion.