Context
This system started as an exploration of how AI tools could be used not just for coding, but across the entire software development lifecycle.
The goal was to understand how planning, implementation, review, and iteration could be structured when AI becomes part of the workflow.
Problem
Software development is often fragmented across tools, roles, and stages. Even with AI tools, the workflow remains loosely structured.
The problem was to explore whether a more intentional system could bring clarity and consistency to AI-assisted development.
Approach
The system was designed as a structured workflow where different AI tools played specific roles.
ChatGPT was used for design and structuring, while Claude was used for implementation. Human judgment was used to validate, refine, and control outcomes.
What worked
The structured workflow improved clarity in development.
Using different AI tools for different roles helped reduce ambiguity.
What didn't
Coordination between tools required manual intervention.
The system still depends heavily on human judgment.
What changed
This system changed how I think about AI in software engineering — not as a tool, but as part of the workflow design.