I'm Auditing How I Work With AI
I use AI throughout product development, but volume does not prove value, so I am measuring where it helps, where I intervene, and where work gets repeated.
The Blog
Blog about product engineering, software development, AI, and how Dakic ships real software weekly.
What We Still Do Manually When Building With AI
AI can handle more implementation work, but people still need to choose the problem, judge the experience, manage risk, and accept the result.
When AI Gets Faster, Verification Becomes the Bottleneck
Faster AI can produce changes before a team can understand them, so the delivery advantage comes from tighter tasks and stronger evidence.
The Next Release Should Answer a Question
Before a meaningful release goes live, decide what question it should answer, what evidence matters, and what you will do next.
A Backlog Is Not a Plan
A backlog preserves possibilities. A plan makes a choice about what changes now and what the team expects to learn.
What Changed This Week? A Better Question for Product Teams
The most useful weekly update does not list activity. It shows what is newly possible because the team made something real.