Coding is cheap, Software is not! - Agentic Engineering explained

AI assistants write code faster than ever. But faster code is not better software. That distinction is exactly where a trend ends and a real paradigm shift begins.

What this talk is about
Agentic Engineering means using AI agents deliberately across the entire software engineering process — not as autocomplete on steroids, but as a structured, verifiable approach to building software. We look at how autonomous agents take on tasks, make decisions, and integrate into existing development workflows — without sacrificing control over architecture, quality, and maintainability.
Vibe Coding vs. Agentic Engineering
Vibe Coding — the approach of just prompting an AI and seeing what comes out — produces results. Sometimes impressive ones. But it does not produce reliable, scalable software. We break down where the approach works, where it fails, and why the distinction matters for professional software development.

What you'll take away

• A clear mental model for Agentic Engineering
• A sharp line between structured AI use and uncontrolled AI output
• Concrete real-world scenarios
• A foundation for discussion within your own team

This talk is for anyone involved in software — whether you write code daily, design architectures, or lead engineering teams.