Build with AI. Show your own thinking.

AI Code Design editorial · September 2026 · 3 min read

Make your role unambiguous

A portfolio should help someone understand what you contributed. Say where you used AI for drafts, implementation, or exploration, and explain what you directed, reviewed, and changed. Tool use does not remove the need for an honest account of your work.

Tell the story of a decision

Choose a moment where you had to make a tradeoff. Perhaps you reduced the number of steps, changed the content order, or removed a feature after testing. Show the alternatives, the evidence available, and the reason for the choice. A decision is more informative than a list of software names.

Separate outcomes from hopes

If the project is a concept, label it as a concept. If you tested with a few people, explain the scope instead of presenting it as broad validation. Avoid invented conversion lifts, testimonials, or adoption numbers. What you learned from a limited test can still be valuable.

Show the work in a readable form

Use focused screenshots with captions, a safe working demo, and a short explanation of the implementation when relevant. Do not force a reviewer to read an entire AI conversation to find your contribution. Make the problem and the resulting experience easy to understand.

Close with a credible next step

Explain the remaining limitation and what you would investigate next. A thoughtful boundary demonstrates awareness of the difference between a polished prototype and a maintained product. The best closing section makes your judgment visible while leaving room for continued learning.