Amazon Leadership Principles — Behavioral Stories (STAR)

Coaching an intern to use AI more effectively

• Amazon LPs — Hire and Develop the Best · Earn Trust · Dive Deep · Insist on the Highest Standards
• Situation — I was the onboarding buddy and code reviewer for a software engineering intern, regularly reviewing his PRs. I noticed a pattern: implementations solved the happy path but repeatedly missed edge cases, didn't follow our coding conventions, and sometimes used patterns inconsistent with our codebase.
• Situation — At first I treated each PR independently with review comments, but after the same issues recurred, I realized the problem wasn't a specific feature — it was how he approached development.
• Task — Rather than leaving the same comments on every PR, I wanted to help him build better engineering habits so he could become more independent and submit higher-quality code.
• Action — Instead of assuming the cause, I scheduled a 1:1 to understand his workflow. I learned he relied heavily on AI tools to generate most implementations — fast, but often submitted without fully understanding the code or validating edge cases.
• Action — I made clear I wasn't against AI — I use it myself. I shared how I integrate it: to brainstorm, clarify concepts, or draft, but always review, test, and adapt before committing. I walked through examples from his own PRs, showing why some AI output didn't fit our architecture or missed business scenarios, and how to think critically about suggestions.
• Result — Over the following weeks his PRs improved in quality and consistency — he began catching edge cases himself, aligned with our standards, and needed fewer repeated comments.
• What I Learned — Mentoring works better when you address the underlying behavior, not just the symptoms. Coaching him to use AI responsibly — rather than discouraging it — made him a stronger engineer while preserving the productivity benefits.