Beyond Clean Code: Refactoring for Maintainable Software
While AI accelerators make code generation trivial, software maintainability, system design, and long-term quality remain critical engineering responsibilities. This session delivers a practical framework bridging code refactoring, software architecture, and AI-assisted engineering. Participants will learn to evaluate AI-generated code, reduce cyclomatic and cognitive complexity, enforce decoupling and cohesion, and safely refactor living systems. Moving beyond superficial clean-code dogma, we cover dependency management, architectural fitness functions, and strategies to prevent architectural erosion in growing applications. Key Takeaways & Core Concepts: - AI Code Auditing & Refactoring: Trace execution, spot code smells (God objects, tight coupling, primitive obsession), and refactor AI output safely. - Complexity & System Boundaries: Master local reasoning, manage accidental complexity, and establish maintainable extension points. - Architecture & Dependency Inversion: Apply SOLID principles, layered patterns (Hexagonal/Clean Architecture), and manage dependency direction. - Safe System Modernization: Use characterization testing, Strangler Fig patterns, and feature flags to refactor without regressions. - Automated Guardrails: Implement architecture tests and continuous refactoring workflows to stop technical debt from compounding.
While AI accelerators make code generation trivial, software maintainability, system design, and long-term quality remain critical engineering responsibilities. This session delivers a practical framework bridging code refactoring, software architecture, and AI-assisted engineering. Participants will learn to evaluate AI-generated code, reduce cyclomatic and cognitive complexity, enforce decoupling and cohesion, and safely refactor living systems. Moving beyond superficial clean-code dogma, we cover dependency management, architectural fitness functions, and strategies to prevent architectural erosion in growing applications. Key Takeaways & Core Concepts: - AI Code Auditing & Refactoring: Trace execution, spot code smells (God objects, tight coupling, primitive obsession), and refactor AI output safely. - Complexity & System Boundaries: Master local reasoning, manage accidental complexity, and establish maintainable extension points. - Architecture & Dependency Inversion: Apply SOLID principles, layered patterns (Hexagonal/Clean Architecture), and manage dependency direction. - Safe System Modernization: Use characterization testing, Strangler Fig patterns, and feature flags to refactor without regressions. - Automated Guardrails: Implement architecture tests and continuous refactoring workflows to stop technical debt from compounding.