Applying engineering to Vibe Coding
Vibe Engineering
How to combine the power of AI agents with software engineering principles to produce higher-quality code faster.
“Vibe Engineering = Vibe Coding + Software Engineering
”
Workflow
The iterative Vibe Engineering cycle: assign the task to the agent, grab a coffee, and review the Pull Request.
Context Engineering
What forms part of an AI agent's context and why it is crucial to configure it correctly.
Talk contents
Five blocks connecting Vibe Coding with professional software engineering.
Introduction to Vibe Engineering
What Vibe Coding is, why it is not enough on its own, and how Vibe Engineering combines AI with software engineering best practices.
Where to apply it?
When to use and when to avoid AI agents: mechanical tasks, refactors, and bugs versus business logic, data models, and interfaces.
Context Engineering
Setting up the information agents need: the master document (AGENTS.md), code files, project structure, and terminal output.
Testing
The test pyramid adapted for Vibe Engineering: more unit tests, TDD as the ideal process for agents, and regression protection.
Conclusions
Vibe Coding is here to stay. The developer role evolves toward product manager, achieving higher-quality code in less time.