How AI Has Changed an Engineering Leader’s Role

Summary

An engineering leader says AI has made hands-on technical work more valuable alongside management. Using Claude, the leader built a simple feature-flagging system after a broken feature took too long to revert; the team adopted the command-line tools for critical deployments, improving deployment quality and reducing risk. AI also lets the leader learn unfamiliar parts of a system quickly by asking for explanations, alternatives, and technical guidance, including on PostgreSQL performance. The team has used AI to improve READMEs and CLAUDE.md files, making it easier for new or newly assigned engineers to locate relevant code, understand it in context, and find the right person to ask. Faster onboarding has made the author less resistant to adding new languages, services, and microservices, although the risks of premature decomposition remain. The author is also less worried about bus-factor problems because people can gather system knowledge faster and AI can examine what code actually does rather than relying only on outdated documentation. The article argues that engineering teams are still developing practical intuitions about AI, while many executives and product managers have not yet formed a clear view of how AI could change products or enable new problem categories. The author expects useful ideas to emerge from people who use these tools daily.