How I make sense of complex backends
How I understand complex backends by identifying system boundaries, making inputs and outputs explicit, and diagramming the data journey.
All the articles I've archived.
How I understand complex backends by identifying system boundaries, making inputs and outputs explicit, and diagramming the data journey.
Reviewers should ask for smaller PRs or more context when a review is difficult, and authors should welcome that pushback.
How distributed behavior, hidden execution paths, and knowledge of several systems made backend code harder to understand coming from frontend development.
Prompt, useful code reviews unblock teammates, reduce downstream rework, and make team delivery a shared priority.
Unblock your teammates ↘
A founder’s personality shapes company habits and structure. Understanding that influence gives you a better starting point for organizational change.
Features can get enterprise software bought. Good UX helps users get value, adopt it, and advocate for it.
UX earns advocates ↘
Use a disposable AI-built prototype to test architectural assumptions, record what you learn, and build the production system in smaller slices.
Give AI responsibility for implementation, verification, and review so your attention stays focused on defining the goal and judging the result.
Use AI to organize large implementations into cohesive PR stacks and automate their upkeep so your team can review them efficiently.
Make it reviewable ↘
AI makes code cheaper to write, but we still need to decide whether the work matters.
Cheap code. Costly choices.