Are your developers AI ready (part 2)? Not without mastering engineering basics

The experience gap

When a seasoned engineer interacts with AI coding tools, they approach the exchange as a technical conversation with an enthusiastic but sometimes misguided junior colleague. They know what questions to ask, can instantly spot conceptual errors and can verify the correctness of the code they receive.

Junior developers, lacking the experiential knowledge, often struggle to evaluate AI suggestions properly. They may accept problematic implementations or spend considerable time trying to adapt unsuitable proposals without recognising fundamental flaws in approach.

This creates a growing divide. Teams with strong engineering practices accelerate while those without them risk amplifying existing problems.

Code quality at speed

Traditional software engineering embraced the adage, “make it work, make it right, make it fast”- a progressive refinement that ensured functionality before optimisation. AI-assisted development compresses this timeline dramatically, particularly the make it work element.

Strong teams leverage AI to iterate through these stages rapidly while still maintaining quality:

• The use AI suggestions to quickly achieve working implementations.

• They refactor with purpose, driven by engineering principles rather than the tools' outputs.

• They focus human creativity on architectural decisions and edge cases.

Without these disciplined boundaries, teams can produce superficially impressive code that harbours hidden technical debt, brittle implementations and subtle bugs that only emerge in production.

Architectural clarity

The best engineering teams maintain clear, documented architectural boundaries that inform how they use AI coding tools. This architectural clarity serves as both guide and guardrail.

Strong architectural understanding helps developers:

• Frame prompts effectively to generate contextually-appropriate code.

• Evaluate AI suggestions against patterns and practices.

• Reject implementations that violate architectural principles.

• Guide the AI towards solutions that align with system design.

Teams lacking this clarity often produce disconnected code fragments that don't integrate well, creating architectural drift and increasing maintenance burden over time.

Testing as verification

Engineering Excellence demands rigorous testing… an area where human expertise remains irreplaceable despite Ai advances.

AI tools can generate tests, but they struggle with:

• Identifying edge cases specific to your business domain.]

• Testing integrations across complex distributed systems.

• Verifying behavioural nuances that may not be explicitly documented.

• Recognising context-specific security and performance concerns.

Experienced engineers know that tests serve multiple purposes: they verify functionality, drive software design, document expected behaviour and prevent regression. AI-generated tests often focus narrowly on the first goal whilst neglecting the others.

Teams with strong testing practices use AI to accelerate test coverage while also maintaining human oversight of test quality and completeness.

Refactoring with purpose

Refactoring (improving code structure without changing behaviour) becomes both more important and more challenging in the AI era.

Engineers with strong fundamentals:

• Recognise code smells that indicate deeper problems.

• Understand when refactoring creates value and when it creates unnecessary churn.

• Follow established refactoring patterns to transform code safely.

• Automate routine refactoring while reserving judgement for complex cases.

AI tools excel at suggesting specific refactoring patterns, but lack the contextual understanding to determine when refactoring is warranted. Without engineering discipline, teams may over-optimise unimportant code whilst missing critical structural issues.

Additionally, because they work by predicting the most likely next word, AIs are prone to introducing subtle, difficult-to-spot defects when performing automated refactorings.

Related Insight

Course: Engineering best practices

Build better software by understanding the disciplines, practices and principles that underpin great code.

Knowledge transfer and mentoring

Perhaps the most overlooked aspect of engineering excellence is its role in knowledge transfer and team growth.

Senior engineers in AI-enabled teams:

• Model effective collaboration with AI tools

• Share prompt engineering techniques that produce better results

• Establish quality standards for AI-assisted code

• Provide contextual knowledge that AI tools can’t capture.

This mentoring role becomes more critical in the AI era, not less. Teams that treat AI as a replacement for senior guidance will find their junior developers struggling to develop the very skills that make AI tools effective.

The path to excellence

Building engineering excellence isn’t a one-time investment. It’s an ongoing commitment to fundamentals that pay increasing dividends as technology evolves.

The companies gaining the most from AI aren’t those with the newest tools… they’re those with the strongest engineering culture.

Organisations serious about maximising the value of AI tools should:

• Invest in training that strengthens core engineering skills.

• Establish clear guidance for using AI within established practices.

• Create feedback loops where AI successes and failures inform process improvements.

• Recognise that strong foundations support faster innovation.

The rise of AI coding tools hasn’t diminished the importance of engineering excellence- it has amplified it. As these tools change code production, the differentiator shifts from ‘who can write code’ to ‘who can consistently deliver high-quality, maintainable software systems’.

In this new landscape, engineering excellence isn’t just a technical consideration: it’s a strategic necessity that determines how effectively organisations can leverage AI’s potential.

The question isn’t whether your teams are using AI tools. It’s whether they have the engineering fundamentals to use them effectively.