The constraints differ far more than the technology does.
Startup: small team, runway pressure, unknown product-market fit, low traffic that might 100x or die.
- Managed services over self-hosted - engineer time is the scarcest resource.
- A modular monolith, not microservices - you cannot staff ten services with six engineers.
- One relational database until it genuinely hurts. Postgres handles far more than people expect.
- Optimise for change speed, since the requirements will change.
- Accept technical debt deliberately, and write down where.
Enterprise: large teams, compliance, existing systems, predictable but large traffic.
- Service boundaries matter, because team coordination is the real bottleneck.
- Compliance and audit are design inputs from day one.
- Integration with existing systems often constrains the design more than any technical factor.
- Standardisation and governance beat local optimisation.
- Multi-year migration paths, not rewrites.
The senior answer: "Premature scaling is as expensive as premature optimisation. The right design for 1,000 users is not a smaller version of the design for 100 million - it is a different design, and planning the seams where it will later split is the actual skill."