Skip to main content

How we use AI

AI speeds up research synthesis

In short: AI is part of how we work at every stage, and you should know exactly where. It speeds up synthesis, production and analysis. People run every interview, make every decision, and review every output before it reaches you.

Why it matters now
9:41100%
Synthesis · 12 interviewsClaude · themes from transcripts

Themes · by how many people said it

Can't compare plans9 of 12

Team pricing unclear7 of 12

Wants to try it first4 of 12

✓A researcher checked every quote

Interviews

Run by people

AI

Sorted into themes

Every quote kept with its source

Why it matters now

57%

of research and design professionals cite faster turnaround on user research projects as a benefit of AI.

Maze, Future of User Research 2025
of developers using AI agents agree the agents have increased their productivity.Stack Overflow Developer Survey 2025
69%
more experiments run by companies using AI in their testing platform, compared with companies that hadn't adopted it.Optimizely, Opal AI Benchmark Report 2025
78.7%

Fast is easy now. Getting it right still takes method, and that's where we've put our effort.

AI-enabled product design

The stages haven't changed. What moves between them has

Instead of drawings that someone later has to interpret and rebuild, you get something people can click, on brand and in real code, within days. That makes it cheap to try several ideas, and the version you tested is the version that ships.

✦ AI helps Testing Data flows

Research

  1. Brief
  2. Research✦

The problem, findings and hypotheses

Design

  1. Design system design
  2. Prototype directions✦
  3. Sign-off

Walkthrough of each direction

Options considered, and why one was chosen

Build

  1. Build and instrument✦
  2. Pull request✦
  3. Review

Team review on a shared link

What was built, and the events wired in

Release

  1. Sandbox
  2. Staging

QA, accessibility and user testing

Review notes, QA and accessibility results

Live

  1. Production
  2. Versions side by side
  3. Analysis✦

Behaviour in real traffic

Behaviour and conversion data, per version

Knowledge base

Every record, yours to keep, read by the next brief

AI and people

Where we use AI, and where we don't

AI helps with: Sorting and summarising research material

People: Decide which problem is worth solving

AI helps with: Producing design options and code

People: Talk to your users

AI helps with: Creating test variations

People: Judge which direction is right

AI helps with: Analysing behavioural and test data

People: Review every output

AI helps with: Keeping the project record up to date

People: Answer for what ships

What that means for you: the slow, repetitive parts take less time. We put that time into more research, more testing and more evidence, not fewer hours.

If you have rules about AI, or obligations to your own customers, tell us. We'll work within them, including doing specific work without AI where you need it.

AI FAQs

AI helps us sort research, produce design options and code, and analyse data. People run every interview, make every decision and review every output.

How we use AI

No. It means more of your budget goes into research, testing and evidence, and less into slow production work.

How we use AI

We work in your tools and accounts wherever you prefer, and we agree how your material is handled at the start of each project. We handle personal information under the Australian Privacy Act, and the EU and UK GDPR where they apply.

Standards, data and your IP

Where to next

Last updated

Book a call

Let’s talk about your product

Book a call. We’ll talk about where your product is, what’s getting in the way, and whether we’re the right fit.

Book a call

Or email ben@antiphono.com