From the observatory to the boardroom: Part 3

If you’re a senior manager or executive watching your teams adopt AI, here’s what this 7-day build and 7 years of digital transformation, taught me about what you actually need to do.

Stop waiting for perfect - don’t confuse speed with strategy

The organisations falling behind aren’t the ones moving too fast. They’re the ones waiting for complete certainty before they act. But the organisations creating the most risk right now are the ones moving fast without architectural foundations.

I’ve spent the last decade of my career modernising legacy systems. Systems build by smart people, moving quickly, with the best intentions. The common thread was always the same: decisions made at speed, without governance, without standards, without a view of what the system would need to become.

AI-assisted development is democratising the ability to build. This is genuinely exciting. It’s also genuinely dangerous if your organisation doesn’t have the structures in place to ensure that what is built today doesn’t become the problem you’re paying to fix in 5 years.

Invest in capability, before you invest in tools

The reason I could build effectively with AI is because I understood the domain. The architecture, security, data modelling and system design. AI amplified that knowledge. It cannot substitute for its absence. Your people need a foundation to evaluate AI output, not just generate it.

Redesign the workflow, not just the tooling

McKinsey’s 2025 State of AI report identified workflow redesign as the single most important factor for realising value from generative AI. Adding AI to a broken process gives you a faster broken process. The same applies to development. Adding AI to an undisciplined engineering culture gives you undisciplined engineering at speed.

Establish your architectural standards now, before the tools proliferate.

This is the lesson legacy modernisation teaches repeatedly. The cost of fixing a bad architecture decision grows exponentially over time. The earlier you establish the standards, the data models, the security patterns, the review processes - the cheaper it is to maintain them.

Put governance at the top of your company’s priority list

CEO oversight of AI is the element most correlated with bottom-line impact. This is not a technology decision delegated to IT. It is a strategic, financial, and ethical leadership responsibility. The technical debt your teams are accumulating today will appear on your balance sheet eventually.

That’s the thread connecting everything I work on - courage in leadership, readiness in action. Building systems, and teams, and organisations that the future can trust.