EP9: Designing AI for Humans: How Nile Builds Trust-Centred, AI-Native Organisations

From Eye-Tracking to AI: Two Decades of Understanding Human Behaviour

Nile's roots go back to being one of the first companies to bring eye-tracking technology to the UK, giving the business a long history of deep, evidence-based consumer understanding. Over the years, Nile has guided organisations through major shifts — from early online banking, to citizen-driven banknote design, to insurance marketplace transformation — and now applies those same strategic human-centred design principles to AI transformation.

Neil describes a consistent lesson learned across each technological shift, from desktop usability testing to the rise of smartphones: organisations must stay "one step removed" from the technology itself, and focus first on the underlying human need it's meant to serve — with the "how" coming second.

A Data-Driven Approach to AI Transformation

Neil shares a detailed case study involving a large wealth management organisation with around £15 billion in assets under management. Rather than relying purely on traditional service design mapping — which can be time-consuming for complex organisations — Nile combined two approaches:

1. A rigorous, quantitative analysis of existing processes — enumerating tasks, timings, and the people involved — to prioritise where AI could deliver the greatest benefit.
2. A human-centred assessment of what it actually feels like to operationally deliver those processes day-to-day.

This combined approach identified concrete opportunities — such as reducing onboarding steps by 10x and servicing steps by 4x — while preserving the human value that clients associate with receiving trusted financial advice.

What "AI-Native by Design" Really Means

For Nile, being AI-native isn't just about making an existing process faster — it's about fundamentally rethinking how a task is done to reduce friction. Neil illustrates this with a compelling example from another financial services project: an AI tool designed to speed up writing compliant marketing copy did technically reduce the effort involved, but because users had to sit and wait several minutes for the tool to generate output, the experience felt slower — even though it was objectively far more efficient. It's a clear reminder that measurable efficiency gains don't always translate into a better human experience, and that unintended consequences must be actively designed around.

Nile's Approach to AI Transformation

Neil outlines the core principles guiding Nile's transformation methodology:

• Start with outcomes, not tools — Conversations about AI transformation often get pulled toward flashy new technology and capabilities. Nile insists on first clearly defining the organisational and human-centred outcomes being pursued.
• Deep regulatory expertise — Specialising in regulated industries like financial services, Nile brings practical understanding of frameworks like Consumer Duty, ensuring that AI-driven change delivers genuinely good outcomes for customers rather than simply ticking compliance boxes.
• Flexibility over rigid methodology — AI transformation projects frequently encounter unexpected technical or organisational blockers. Nile deliberately avoids prescriptive, milestone-locked methodologies in favour of an adaptable, pivot-ready approach.

The Three Biggest Mistakes Organisations Make with AI

Drawing on years of transformation work, Neil identifies three recurring pitfalls:

1. Focusing on "how" before "what" — Jumping straight to implementation without clearly defining the actual outcome being pursued.
2. Starting in the wrong place — Failing to use a rigorous, evidence-based (both quantitative and human) approach to choose where to begin a proof of concept.
3. Ignoring the human side of adoption — Overlooking how a new AI tool fits into existing workflows, and what will genuinely drive employee adoption, rather than simply focusing on the technology itself.

Why Human Collaboration Still Matters in an AI-Powered World

One of the most striking reflections in the conversation comes from a recent innovation sprint Nile ran with a client. Despite having access to a full suite of AI tools for idea generation, critique, and synthesis, the team found that having people physically together in a room produced significantly more alignment and collective confidence than relying on AI alone. Neil is careful to note this isn't an argument against using AI tools — Nile's team did use AI-assisted prototyping during the same sprint — but a reminder that certain human-driven practices remain genuinely valuable, and shouldn't be discarded prematurely.

Looking Ahead: Agents, Culture, and Trust

Neil points to three key themes that will have the biggest impact on AI and human collaboration over the coming years:

• AI agents — While transformative, Neil expects adoption to be slower in regulated, risk-averse industries, where the appropriate level of autonomy given to an agent requires careful consideration.
• Organisational culture and values — Rather than being purely technology-driven, decisions about how AI is applied will increasingly be shaped by an organisation's underlying culture and values.
• Consumer trust — As AI reshapes everyday digital experiences, from online shopping to travel booking, trust in which brands and platforms are given access to personal data will become an increasingly critical differentiator.