Blog / Product Engineering

Product Engineering

Taking a product from a real problem to shipped software: scoping, architecture decisions and building the first version yourself.

Kilele Archade and Apex both started as answers to problems I could point at rather than ideas I wanted to try. That constraint changes the whole build: scope gets cut to what a real user needs this month, and the boring choices — one database, one hosting story, scripts for anything done twice — win by default.

These pieces cover the part of the job that is not typing code: deciding what the first version is, keeping solo delivery manageable, and letting actual users move the product further than any technical decision did.

Writing on product engineering (1)

Built in practice

Hire for this

Experiments in the lab

Other topics

← All writing · Work with Fredrick Onyango