Home/How we work
Start with people. Build what matters.
Great digital products start with the right questions about people, their needs and the problems they're trying to solve.
Built for how products get made today.
Good products aren't created by moving neatly from one step to the next. Research can change the strategy, testing can change the design, and AI now makes every loop faster, without cutting corners.
AI does the heavy lifting. People make the decisions.
Our methodology
We use a human-centred design methodology, with Human-led AI built in, to create, transform and improve digital products.
We start with the people, problems and context behind a product, then add business goals and technology to decide what to build. AI drafts the plans, requirements, code and tests. Experienced people question it, correct it and decide, with you, all the way through the project engagement.
The loop
You choose the destination.
You approve every turn.
Usually a person asks AI for help, one task at a time. We turn that around: AI plans the route and people approve every turn, so you stay in control.
AI drafts the plan
Given the goal, AI proposes how to get there: the steps, the options and the risks, written down where everyone can see them.
AI asks the questions
Who is this for? Which outcome matters most? The gaps in a brief surface in the first week, while they are still easy to close.
People decide
Our senior team, with yours, reviews what has been proposed, corrects it and chooses. Nothing moves on until a person has said yes.
AI does the work
With the decision made, AI produces the research summary, the design options, the code or the tests, and the loop starts again.
Our approach
Five stages.
Understand, Define, Design, Build, Evolve. The loop runs inside every stage, with validation and testing woven in rather than saved for the end. Stages overlap and repeat, and some engagements spend most of their time in one or two.
01 Understand
We start with questions,
not assumptions
We research the problem with your team, your data, your users and your market, asking the questions most projects skip. Interviews, workshops and an honest look at what exists mean later decisions rest on evidence, not opinion.
- AI does
- Reads the existing material and pulls patterns out of interviews and data.
- People decide
- What the evidence means and which problem is worth solving.
- You
- Share your goal and the inside knowledge only you have.
The resultA shared, accurate picture of the problem, before anyone starts solving it.
02 Define
We agree what we're actually solving for
We turn what we've learned into a small number of clear priorities and a direction everyone can get behind. Scope gets honest here: we say no to what doesn't matter yet, so what does gets real attention.
- AI does
- Drafts the requirements, the risks and a first plan from your goal.
- People decide
- Which priorities matter and what stays out of scope.
- You
- Review the first draft and shape it before anything is locked in.
The resultA defined problem, a clear direction, and a plan everyone has agreed to.
03 Design
We design for the experience
We take ideas from early concepts through flows, wireframes and prototypes, exploring and refining how the product works and feels. UX and UI design bring clarity, consistency and purpose to every interaction.
- AI does
- Produces more options, faster: alternative flows, working prototypes with realistic content, and checks against the design system and accessibility standards.
- People decide
- Which direction is right. Taste, judgement and what we hear from users in testing stay with experienced designers.
- You
- See real, clickable designs early and sign off the direction before build begins.
The resultA clear, considered experience, ready to move forward.
↓
Validate & test
Before anything expensive gets built, we pressure-test the idea with prototypes and with the people who'll actually use it. Whatever doesn't hold up gets fixed here, while it still costs little to change.
04 Build
We turn design into working products
We work with technology teams to turn the approved design into a working product, in short cycles that each end in something reviewed. You see progress as it happens.
- AI does
- Proposes the technical design, writes the code, and writes and runs the functional, security and performance tests against standards set once and applied every time.
- People decide
- The architecture and the trade-offs. Senior people review every piece of work before it is accepted.
- You
- See progress as it happens and make the calls that matter, with no big reveal at the end.
The resultA product being built the way you expected, with no surprises at the end.
↓
Validate & test
Once it's built, we test it again: functionally, with real users, against real data, and fix what only shows up once something is real.
05 Evolve
We keep improving, making it better
A launch is a milestone, not a finish line. We track how the product is used and where it falls short, and keep refining it against real behaviour.
- AI does
- Watches performance and usage, flags problems early and proposes fixes and improvements.
- People decide
- What to change, what to leave alone and what to do next.
- You
- Get evidence of what is working, and a product that responds quickly to your market and your feedback.
The resultA product that keeps evolving and improving, backed by evidence instead of guesswork.
What it means for you
Better products,
in less time.
Faster delivery
Weeks of back-and-forth become days. The time comes out of hand-offs, waiting and rework, and stays in the thinking.
A closer fit
AI keeps asking questions, so what gets built stays close to what your business and your customers need.
Consistent quality
Your standards are written down once and applied every time, backed by far more automated testing than a team could write by hand.
Quicker to adapt
Short cycles mean we can respond to the market, and to your feedback, while it still matters.
Sized to fit
your project.
No project is too small or too large, but our process is never one-size-fits-all. A tweak or a fix is one short cycle. A redesign across many systems gets all five stages. A new product starts from your goal, and an existing one starts with AI mapping what is there and our team checking the map.
The goal is to create an unforgettable digital experience that creates brand loyalty.
Working together
From first conversation to delivery.
A look at how an engagement actually runs, step by step.
Get in touch
Tell us what you're working on, what isn't working, or where you think the opportunity is. No obligation, no generic pitch.
Scope & proposal
We work out what needs to happen and put it in a proposal, with a cost we agree together. Once it's signed off, a deposit secures the work and the project begins.
Understand & define
We research the problem, agree the priorities and the plan, and open the shared project record, so you know what's happening, when, and what we need from you along the way.
Design & approval
We design and prototype the experience, test it, and bring it to you to review. Nothing moves into build until you've seen it and signed it off.
Build, with you in the loop
We build in short cycles with regular updates, involving you in the decisions that matter rather than just the final reveal. A milestone payment falls around the midpoint.
Handover
We hand over the finished product and its project record, with the final payment due on delivery. If there's an Evolve stage, the relationship continues from there.
Next steps
Digital products people value.
Whether you're creating something new, transforming an existing product or looking for an experienced partner to improve what you already have, let's talk.