Q&A with CCO André Hageraats: Why the UK, why now, and why AI is a stress test for architecture

The UK has no shortage of AI ambition. Organisations are under pressure to improve productivity, reduce costs, support employees, and deliver better customer experiences.

According to the UK Government’s AI Adoption Research, only 16% of UK businesses currently use at least one AI technology. A further 5% plan to adopt it, while the majority (80%) are neither using nor planning to adopt AI.

That hesitation is not just about AI tools. It is about whether organisations have the systems, data, APIs, and governance needed to use AI safely at scale.

As Yenlo increases its focus on the UK market, we spoke with André Hageraats, CCO of Yenlo, about why AI readiness starts with stronger digital foundations.

Why is AI readiness such a challenge for UK organisations?

Many organisations in the UK have spent years investing in cloud platforms, SaaS applications, customer portals, APIs, data platforms, and automation.

Those investments helped them modernise. But they also created complexity.

Systems are often connected in different ways. APIs may be governed inconsistently. Critical data can sit across multiple platforms. No single team may have a complete view of how everything connects.

For AI, that matters.

If systems are disconnected, AI only sees fragments. If data quality is inconsistent, AI outputs will be inconsistent. If access controls are unclear, AI can create compliance, ethical, or security risks.

AI does exposes architectural complexity.

What is the biggest misconception about scaling AI?

Many organisations start with the question:

“Which AI tool should we use?”

A better question is:

“Can we give AI the right information, from the right systems, with the right controls, at the right time?”

An AI proof of concept can work with manually prepared data and a narrow use case. Production AI is different. It needs reliable integrations, trusted data, clear ownership, strong access controls, and governance.

Without those foundations, AI is difficult to scale safely.

Can you give an example?

Take logistics.

A company may want to use AI to optimise delivery routes, reduce fuel costs, improve warehouse planning, and provide more accurate delivery updates.

But the AI may need vehicle telemetry, warehouse capacity, driver schedules, order data, customer SLAs, traffic feeds, inventory information, and exception data from operational systems.

If those systems are not connected properly, the AI only sees part of the picture.

It may optimise a route without understanding warehouse cut-off times. It may prioritise a delivery without knowing the customer’s contractual SLA. It may recommend a plan that looks efficient but does not reflect real operational constraints.

The same applies in financial services. A generative AI assistant may need information from CRM, policy systems, transaction platforms, document repositories, and legacy databases. If those systems are not governed and connected, the assistant may give incomplete answers or expose sensitive information.

AI increases the need for integration that is secure, governed, and reliable.

Where do WSO2 and Boomi fit in?

Different organisations need different integration strategies.

WSO2 is often a strong fit for organisations that need deeper control over APIs, identity, security, governance, and enterprise-grade integration architecture.

Boomi is often a strong fit where speed, simplicity, and rapid connectivity across SaaS applications are priorities.

The platform should follow the strategy, not the other way around.

Yenlo helps organisations understand the current landscape, identify risk and duplication, clarify governance needs, and design an integration strategy that supports long-term growth.

Where should organisations start?

Start with visibility.

Before scaling AI, organisations need to understand which systems matter, where critical data lives, how integrations are governed, and where access or ownership is unclear.

That means mapping the integration landscape, reviewing API governance, identifying critical dependencies, and assessing whether existing systems are secure and reliable enough to support AI use cases.

AI rewards organisations that have control over their architecture. It exposes those that do not.

The organisations that generate the most value from AI will not necessarily be the ones that spend the most on AI tools. They will be the ones that can connect, govern, secure, and trust the information AI depends on.