How I choose backend technology without paralysis

For a long time I searched for the “perfect” backend technology. The greatest cost did not come from a wrong choice; it came from the times I chose nothing.

Context first

Today I start with the problem: where risk lives, which constraints are real, who will operate the system, and which decisions are reversible. Languages and frameworks matter, but they follow foundations such as HTTP, data, observability, and clear boundaries.

My practical rule

I pick an option good enough to learn from real use. I record why, define signals that would justify revisiting it, and move. A technical decision is not a marriage; it is a verifiable commitment to the current context.