AI is part of our everyday toolkit, and we think clients deserve to know exactly how. Our rule of thumb is simple: AI speeds up the work around decisions; people make the decisions. Here is what that looks like at each stage of a project.
Research#
Research is where the temptation to cut corners is greatest, and where it matters most not to. AI can make the work around research faster. It cannot replace watching real people.
AI helps with
- Organizing notes from many sessions
- Suggesting first-pass themes to check
- Drafting interview guides and task wording
People decide
- Running every session and watching real behavior
- Deciding which patterns actually matter
- Checking every theme against the evidence
Any themes AI suggests are a starting point for checking, never a finding in themselves.
Design#
In design, AI is useful for breadth: exploring more options, more quickly, so the good ones are easier to find. Choosing between them is still a matter of judgment, taste and knowing the people the design is for.
AI helps with
- Exploring layout and wording variations
- Drafting first versions of content
- Suggesting image descriptions to refine
People decide
- Which direction to take, and why
- Final wording, tone and accessibility
- Whether it works for real people, tested with them
Build#
AI coding tools save time on repetitive work and are good at spotting problems. Every line that ships is still reviewed by a person who understands it.
AI helps with
- Routine code and first drafts
- Writing tests and finding bugs
- Second-opinion code reviews
People decide
- Structure, security and performance
- Reviewing everything before it ships
- What goes live, and when
The lines we don’t cross#
- No client data in AI tools without permission. Research recordings, customer information and private documents stay private unless you agree otherwise.
- No simulated users instead of real ones. AI can help us organize what real people said. It does not stand in for them.
- No unreviewed output. Nothing written, designed or built by AI reaches you, or your customers, without a person checking it.
- No unchecked claims. Facts, figures and sources are verified, however confident an AI answer sounds.
What this means for clients#
In practice, AI lets us explore more options and spend more of our time on the parts of a project that need human judgment: understanding your customers, making the hard calls and getting the details right. The accountability doesn’t change. If AI played a meaningful part in your project, we will tell you where and how.