AI UX design course in New Zealand. Design with AI. Lead with people.
- Length
- 8 weeks, part-timeAbout 10 hours a week
- Price
- $4,789 + GSTFounding cohort $3,997 + GST
- Format
- Online, livePlus the IDNZ learning portal
- Mentoring
- Unlimited one-to-oneMon–Fri, plus weekend slots
- Tools
- Included for 8 weeksFigma, Claude, ChatGPT
- Leads to
- CAAXDE certificationCounts towards CAUXPDE
What AX means
AI and agent experience design, in three parts.
UX is how people experience a product. AX adds two new users of your work: the AI inside the product, and the AI agents that use it. CAAXDE teaches all three parts below, on top of the research and testing you already do.
AI features people can trust
What an AI feature suggests, how it explains itself, and how people check it, correct it or switch it off.
For exampleA booking app that suggests a time, says why, and lets you change it in one tap.
Products AI agents can use
Sites and apps that an AI agent acting for a person can read and complete tasks on, without making things worse for the people who use them.
For exampleA council service page an agent can read to book an inspection, with the same clear steps a person sees.
Building with AI agents
Handing your Figma work to Claude Code or Codex through Figma’s MCP server to build a prototype you can test, then checking every result yourself.
For exampleA tested, clickable prototype in an afternoon, built from your own design system.
AI is the icing. The research and the decisions are yours.
Who it’s for
For designers who already do UX.
Working UX, UI and product designers
You design for a living and need to be current with AI in 2026 and 2027: what to use it for, what to keep for yourself, and how to design products that contain it.
CAUXPDE and other UX graduates
You know the research-to-prototype process and want the AI and agent layer on top. CAUXPDE graduates get a step-up price.
Product teams
Your employer wants AI used well, with research and testing behind it. Teams of three or more get a lower rate, and eight or more can run a private cohort.
New to UX? CAAXDE isn’t a beginner course. Start with the 5-week Agile UX Design JumpStart, then the 11-week CAUXPDE, then come back for CAAXDE.
What you need: UX, UI or product design experience, or a completed UX course; a laptop that runs Figma; and about 10 hours a week for 8 weeks. You don’t need to know how to code.
Try AX design
Four switches between an AI guess and a feature people trust.
This is the kind of AI feature you design in week 6. Turn each pattern on and watch the same suggestion become something people can recognise, check and control. In the course you test it with real users, then write the spec.
Patterns based on Microsoft’s 18 guidelines for human–AI interaction, taught in week 6.
Kōwhai Health
Book a GP visit
Thu 10:30am
Dr Ngata
Why this time? You’ve booked Thursday mornings with Dr Ngata three times.
Book Thu 10:30am
Turn off suggestions
of 4 trust patterns on
Switch them all on.That’s a feature people can trust.
Always current
AI design changes every few weeks. CAAXDE changes every two months.
You learn what’s working in AI design now, and what’s coming next. Every two months the course is reviewed and updated, with a dated note of what changed, so new tools reach you within weeks, not months.
6
course updates a year, each with a dated change note
Every two months, from the first intake
2 months
at most between a major tool release and the next update
The longest a change waits for review
9
changes designers had to learn in 10 months
Nov 2024 to Aug 2025, listed below with sources
Nine changes in ten months · Nov 2024 to Aug 2025Tap a point
Nov 2024
Anthropic introduces MCP
The open standard that later lets AI agents read your Figma files.
CAAXDE Week 5
Source: Anthropic (opens in a new tab)
Jan 2025
NZ Public Service AI Framework
The rules NZ government agencies follow for safe, transparent AI.
CAAXDE Week 8
Feb 2025
Claude Code research preview
An AI agent that writes and runs code from plain-language instructions.
CAAXDE Week 5
Source: Anthropic (opens in a new tab)
May 2025
Figma Make announced
Prompt-to-prototype, inside the tool designers already use.
CAAXDE Week 4
Source: Figma (opens in a new tab)
May 2025
OpenAI launches Codex
A second coding agent, built into ChatGPT.
CAAXDE Week 5
Source: OpenAI (opens in a new tab)
May 2025
Claude Code generally available
Agent builds move from preview to an everyday tool.
CAAXDE Week 5
Source: Anthropic (opens in a new tab)
Jun 2025
Figma’s MCP server (beta)
AI agents can read your design system: components, variables and styles.
CAAXDE Week 5
Jul 2025
NZ’s first national AI strategy
Government backs AI adoption across New Zealand businesses.
CAAXDE Week 8
Source: MBIE (opens in a new tab)
Aug 2025
GPT-5 released
A new default model in ChatGPT, with new behaviour to design and test around.
CAAXDE Week 7
Source: OpenAI (opens in a new tab)
How long a change takes to reach learners
From a change in practice to teaching it. Bars drawn to scale, in months.
For a new degree programme, NZQA asks providers to give at least three months’ notice before applying, and aims to complete approval within six months of the application (NZQA, opens in a new tab). That process is what gives you a formal NZQA qualification, and it suits people who need one. CAAXDE is an IDNZ industry certification, not an NZQA qualification, so it can change as fast as the tools do.
What CAAXDE covers
Eleven things in the syllabus, each one practised.
Use this list to compare AI UX courses. Every topic is taught live, practised in the tool studio and handed in as part of your project, with design thinking running through all of it. Pick a topic to see what you practise, then roll over the picture to see where AI helps and what stays yours.
Design thinking · how the project runs
Design thinking, from real people to a tested prototype
Design thinking is built into CAAXDE. Your project runs through all five stages, each in its own week, and at every stage you learn where AI speeds things up and where your judgement has to lead.
- Empathise: five real interviews in week 2
- Define: an opportunity solution tree in week 2
- Ideate and prototype: flows, Figma Make and agent builds, weeks 3 to 5
- Test: five users and AI evals in week 7
Taught inWeeks 1 to 7
Weeks 1 to 7 · the whole project Where AI helps, and what stays yours
- AI helps with
- Sorting interview notes into themes
- First ideas and drafts to react to
- Fast prototypes you can put in front of people
- You decide
- Who you talk to, and what you ask
- Which problem is worth solving
- What goes in front of real users
You hand inEvery stage shows up in your weekly hand-ins and your final case study.
Prompting · Claude and ChatGPT
A seven-part prompt framework for design work
Briefing Claude and ChatGPT so the output is usable the first time, and knowing where AI shouldn’t be used in UX work at all.
- The seven-part prompt framework
- Claude and ChatGPT projects set up for your brief
- Where AI helps, and where it doesn’t belong
- A prompt library you keep using
Taught inWeek 1
Week 1 Where AI helps, and what stays yours
- AI helps with
- First drafts of copy, flows and research plans
- Variations to compare side by side
- Summaries of long documents
- You decide
- The context and constraints in the brief
- What’s usable and what isn’t
- When to leave AI out
You hand inYour project brief and a prompt library you’ll keep using.
Research · with real people
Research with five real people before any AI prototyping
Interview guides, five interviews and evidence-based synthesis. AI helps you sort what people said; it never invents what they didn’t.
- Writing an interview guide
- Five real interviews
- Evidence-based synthesis
- An opportunity solution tree
Taught inWeek 2
Week 2 · mentor check-in Where AI helps, and what stays yours
- AI helps with
- Transcripts of your interviews
- Sorting notes into first themes
- Spotting quotes for each theme
- You decide
- The interview guide and the five interviews
- Checking every theme against what people said
- The opportunities worth chasing
You hand inResearch synthesis and your opportunity solution tree.
Accessibility · WCAG 2.2 AA
Accessibility to WCAG 2.2 AA, across all eight states
Card sorts, tree tests and flows that include errors, empty and loading screens, checked against WCAG 2.2 AA so the product works for everyone.
- Card sorting and tree testing
- Flows that cover all eight states
- Error, empty and loading screens
- WCAG 2.2 AA checks
Taught inWeek 3
Week 3 Where AI helps, and what stays yours
- AI helps with
- First-pass contrast and alt text checks
- Draft copy for error and empty states
- Spotting missing states in a flow
- You decide
- Testing every state yourself
- Meeting WCAG 2.2 AA
- How the structure is organised
You hand inYour flows and tree-test results.
Design system · Figma and Figma Make
Figma variables, tokens and Figma Make
A design system with variables, tokens and components, then your first Figma Make prototype built from it.
- Figma variables and tokens
- Components and variants
- Your first Figma Make prototype
- A system an agent can build from in week 5
Taught inWeek 4
Week 4 · mentor check-in Where AI helps, and what stays yours
- AI helps with
- Figma Make turning a prompt or a design into a working prototype
- Suggesting component variants
- Filling screens with realistic content
- You decide
- The variables, tokens and components
- What the prototype needs to test
- Fixing what Figma Make gets wrong
You hand inYour design system and first Figma Make prototype.
Agent builds · Figma’s MCP server
Agent builds through Figma’s MCP server
Connecting your Figma file to Claude Code and Codex through Figma’s MCP server, so an agent builds a prototype you can test from your own design system.
- The Lean UX canvas and story mapping
- Figma’s MCP server, Claude Code and Codex
- A prototype built from your design system
- Reviewing every result yourself
Taught inWeek 5
Week 5 Where AI helps, and what stays yours
- AI helps with
- Reading your file through the MCP server
- Building the prototype from your components
- Fixing what you point out
- You decide
- The design system it builds from
- The stories it has to cover
- Checking every screen before testing
You hand inYour story map and an agent-built test prototype.
AI features · trust
Designing AI features people trust
Disclosure, explanation, correction, levels of autonomy and delegation, with Microsoft’s 18 guidelines for human–AI interaction.
- Saying it’s AI
- Explaining why
- Easy correction and an off switch
- How much the AI should do on its own
Taught inWeek 6
Week 6 · mentor check-in Where AI helps, and what stays yours
- AI helps with
- The suggestion or action itself
- A plain-language reason for it
- Learning from people’s corrections
- You decide
- What it says about itself
- How people check, change or switch it off
- How far it’s allowed to act alone
You hand inAn AI feature spec for your project.
AX audit · a real NZ product
Agents as users, and an AX audit
Checking whether AI agents acting for someone can read a real NZ website or app and finish a task on it, without making it worse for people.
- Agents as users of your product
- Can an agent read it and finish a task?
- The trust patterns in its AI features
- What to fix first, with evidence
Taught inWeek 6
Week 6 Where AI helps, and what stays yours
- AI helps with
- An agent attempting a real task on the site
- A log of where it got stuck
- A first list of issues
- You decide
- Which tasks to test
- Judging what failed and why
- What to fix first
You hand inYour AX audit, a second portfolio piece.
AI evals · for designers
AI evals for designers
Golden sets, error analysis and checking an AI grader against people, alongside usability testing with five users.
- Usability testing with five users
- A heuristic review
- Golden sets and error analysis
- Checking an AI grader against people
Taught inWeek 7
Week 7 Where AI helps, and what stays yours
- AI helps with
- Grading answers against your golden set
- Running the same checks every time
- Grouping the errors it finds
- You decide
- The examples in the golden set
- Whether the grader agrees with people
- What the five users show you
You hand inYour test report and eval set.
Governance · NZ and beyond
NZ privacy, EU AI Act disclosure and Māori data sovereignty
How the Privacy Act 2020, EU disclosure rules and Māori data sovereignty shape the AI features you design, then a measurement plan and handoff.
- The Privacy Act 2020
- EU AI disclosure rules
- Māori data sovereignty
- A measurement plan and handoff
Taught inWeek 8
Week 8 · mentor check-in Where AI helps, and what stays yours
- AI helps with
- Drafting disclosure copy to review
- Listing the data a feature touches
- Summarising rules into a checklist
- You decide
- What data the feature should use
- How it’s disclosed to people
- What success looks like, and how it’s measured
You hand inFinal case study, showcase presentation and timed practical.
Always current · every two months
A course updated every two months
Each update gets a dated note of what changed in the learning portal and on this page, and graduates get tool drop sessions for 12 months.
- A review every two months
- A dated change note each time
- New tool features added to the weeks they belong in
- Tool drop sessions for graduates, for 12 months
Taught inEvery 2 months
Every 2 months Where AI helps, and what stays yours
- What changes
- New features in Figma, Claude and ChatGPT
- Changes to how agents read products
- New rules for AI disclosure
- What you get
- A dated note of what changed
- Updated lessons in the learning portal
- Tool drops for 12 months after you finish
You hand inNothing extra. Updates are included.
CAAXDE syllabus, November 2026. Last reviewed . Figma, Claude, ChatGPT and Codex are trademarks of their owners; IDNZ isn’t affiliated with or endorsed by them.
What you leave with
Two portfolio pieces, and the tools to keep going.
Everything you make in the 8 weeks is real work you can show: a case study from a live brief, and an audit of a real New Zealand product. Paid access to the tools is included for the whole course.
A complete case study from a real project
From the problem and five real interviews through to a tested prototype and a spec for its AI feature.
An AX audit of a real NZ website or app
How ready it is for AI agents, and how well it handles disclosure, explanation and correction. A second portfolio piece.
Working use of the tools, including agent builds
Figma, Figma Make, Claude and ChatGPT, and building prototypes with Claude Code or Codex through Figma’s MCP server.
The CAAXDE certification
An IDNZ industry certification, assessed on your work. It counts towards CAUXPDE if you take that next.
12 months of tool drops
Free “tool drop” sessions after major releases of Figma, Claude or ChatGPT, so you don’t fall behind once the course ends.
Tools and learning portalIncluded
Figma Design and FigJam, with Figma’s MCP server
Research boards, flows, variables, tokens and components, and connecting your file to Claude Code and Codex in week 5.
Figma Make
Your first AI-built prototype in week 4.
Claude
Chat with connectors, and Claude Code for agent builds.
ChatGPT
Projects, developer mode and Codex.
IDNZ learning portal
Open 24/7: each module’s lessons, reading material and activities, and where you hand in your work. All included.
Paid access to Figma, Claude and ChatGPT is included for the 8 weeks. The Figma seat is what unlocks its MCP server, so you can use every feature taught without buying a plan.
Figma, Claude and ChatGPT are trademarks of Figma, Inc., Anthropic and OpenAI. IDNZ isn’t affiliated with or endorsed by them.
Week by week
Eight weeks, one real project.
Each week builds the same project a step further, from the brief and real interviews to a tested prototype and its AI feature. Study around your job: about 10 hours a week, online. Open a week to see what you learn and what you hand in.
A typical week
- Learn the week’s module in the learning portal: reading, activities and project work, about 6 hours.
- Practise with the tools in the hands-on studio, 2 hours.
- Book your mentor as often as you need: unlimited Monday to Friday, with limited slots on Saturday and Sunday.
- Hand in the week’s piece of your project and go through it with your mentor.
Each weekAbout 10 hours
Your live lessons are your one-to-one mentor sessions, booked through the calendar we give you. Syllabus: November 2026, the first intake. The course is updated every two months, with a dated note here and in the learning portal of what changed.
00
Week 0Set-up week
Accounts, installs and your briefYou’ll learn
- A tool set-up clinic: Figma, Claude and ChatGPT accounts, with Claude Code and Codex installed
- Your project brief is released
- Your first one-to-one mentor session
You hand in
Nothing yet. You start week 1 ready to go.
In the studioAll of them, set up and working
01
Week 1AI in the design workflow, and the brief
Design thinking, the prompt framework and kick-offYou’ll learn
- Design thinking: the five stages your project follows, and where AI fits in each
- The seven-part prompt framework in Claude and ChatGPT
- Where AI helps in UX work, and where it shouldn’t be used
- Project kick-off with the partner organisation
You hand in
Your project brief and a prompt library you’ll keep using
In the studioClaude and ChatGPT projects
02
Week 2Research with real people
Five interviews and evidence-based synthesisMentor check-inYou’ll learn
- Writing an interview guide
- Five real interviews
- Evidence-based synthesis: AI helps you sort, never invent
- An opportunity solution tree
You hand in
Research synthesis and your opportunity solution tree
In the studioFigJam and Claude
03
Week 3Structure and accessibility
Card sorts, flows, tree tests and WCAGYou’ll learn
- Card sorting
- Flows that cover all eight states, including errors and empty screens
- Tree testing
- WCAG 2.2 AA
You hand in
Your flows and tree-test results
In the studioFigJam and Figma
04
Week 4Interface and design system
Variables, tokens and Figma MakeMentor check-inYou’ll learn
- Figma variables, tokens and components
- Your first prototype in Figma Make
You hand in
Your design system and first Figma Make prototype
In the studioFigma and Figma Make
05
Week 5Lean UX, and building with agents
Story maps and agent-built prototypesYou’ll learn
- The Lean UX canvas, user stories and story mapping
- Connecting Figma to Claude Code and Codex through Figma’s MCP server
- Building a prototype you can test, from your own design system
You hand in
Your story map and an agent-built test prototype
In the studioFigma’s MCP server, Claude Code and Codex
06
Week 6Designing AI features and agents
Trust, autonomy and the AX auditMentor check-inYou’ll learn
- Trust, disclosure and correction
- Levels of autonomy and delegation
- Agents as users of your product
- Microsoft’s 18 guidelines for human–AI interaction
- The AX audit of a real NZ website or app
You hand in
An AI feature spec and your AX audit
In the studioClaude and ChatGPT
07
Week 7Testing and evals
Usability tests and AI evalsYou’ll learn
- Usability testing with five users
- A heuristic review
- AI evals: golden sets, error analysis, and checking an AI grader against people
You hand in
Your test report and eval set
In the studioYour prototype, Claude and ChatGPT
08
Week 8Govern, measure, ship
Privacy, measurement and the showcaseMentor check-inYou’ll learn
- NZ privacy rules, EU AI disclosure and Māori data sovereignty
- A measurement plan and handoff
- Showcase day with the partner panel, and your assessment
You hand in
Final case study, showcase presentation and timed practical
In the studioFigma and FigJam
By week 8 you’ll have
- A prompt library
- Research synthesis and an opportunity solution tree
- Flows and tree-test results
- A design system and a Figma Make prototype
- A story map and an agent-built prototype
- An AI feature spec
- An AX audit of a real NZ product
- A test report and an eval set
- A final case study and showcase presentation
All of it comes from one live brief, so you finish with real work to show alongside your certification.
How it finishes
- Showcase and practicalPresent your case study to the partner panel in week 8, then the timed practical. There’s no 200-question exam.
- Get certifiedYour CAAXDE certification: an IDNZ industry certification, assessed on your work.
- 12 months of tool dropsFree sessions after major releases of Figma, Claude or ChatGPT.
- Step up to CAUXPDECAAXDE counts towards the 11-week CAUXPDE if you take it next.
Real work
A live brief, and a second piece for your portfolio.
You don’t practise on made-up apps. Partner organisations are announced each intake, brief the cohort in week 1 and sit on the showcase panel in week 8.

Weeks 1 to 8
A live brief from a New Zealand partner organisation
You take one product problem from research to a tested prototype and a spec for its AI feature. Already in a product team? Use your employer’s product instead, with their agreement.
- Kick-off with the partner in week 1
- Five real interviews in week 2
- An agent-built prototype in week 5, tested with five users in week 7
- Showcase to the partner panel in week 8

Week 6
An AX audit of a real NZ website or app
You review how ready a real product is for AI agents, and how well it handles the trust patterns people need from AI features. It becomes a second piece for your portfolio.
- Can an agent acting for someone read it and finish a task?
- Does each AI feature say what it is, explain itself and let people correct it?
- What to fix first, with evidence
Mentors who review your AI work, one to one.
Bring your prompt library, an agent-built prototype or an eval set and get specific feedback on it. Book as many one-to-one sessions as you need from week 0 to week 8, with mentors who have worked in New Zealand for more than ten years.
Sai Constantine
Head of Digital, IDNZ
Sai leads agentic UX design at IDNZ and mentors CAAXDE students through the AI and agent side of the course: the prompt framework, Figma Make, building with Claude Code and Codex through Figma’s MCP server, and designing AI features people can trust. Before IDNZ, Sai worked in digital media at Fairfax and led sales and business development at Domainz (Melbourne IT) and MyRepublic. He went through a Creative Technologies degree.
- Agentic UX design
- AI features and agents
- Figma Make
- Claude Code and Codex
- Design systems
- Prompt frameworks
Seann Lautner
UX researcher and mentor, IDNZ
Seann mentors CAAXDE students on the human side of every project: the interview guide and five interviews in week 2, tree tests, usability testing with five users in week 7, and checking an AI grader against real people. Formerly a UX researcher at Air New Zealand and Vodafone NZ, and a Certified Scrum Master and agile coach.
- User research
- Usability testing
- Lean UX
- Evals with real people
- Agile UX
- Scrum
How mentoring works on CAAXDE
Your live lessons are your mentor sessions. Book them through the calendar we give you, as often as you need.
- Week 0First session
- Week 1Open sessions
- Week 2Check-in
- Week 3Open sessions
- Week 4Check-in
- Week 5Open sessions
- Week 6Check-in
- Week 7Open sessions
- Week 8Check-in
- Monday to Friday
- Unlimited sessions
- Saturday and Sunday
- Limited slots
- Project check-ins
- Weeks 2, 4, 6 and 8
In their own words, with the same mentors.
CAAXDE starts on 15 November 2026, so it has no graduates yet. These are UX students Sai and Seann have mentored on IDNZ’s other UX courses.
- 1,000+
- students and business owners trained since 2016
- 5.0
- from 90+ Google reviews
- 86%
- of graduates advance their career within 12 months (IDNZ graduate survey)
- 10+
- years each mentor has worked in New Zealand
“It was an amazing journey, right from the start until the end. The course challenged me to think to a more human-centered design, by learning the agile way with a real project. Thank you so much to my mentor Seann Lautner, who was incredibly patient.”
“I completed the Agile Digital Product Design course at the Institute of Digital Marketing NZ, and it was a genuinely valuable experience. Seann was an excellent mentor—approachable, knowledgeable, and great at breaking down complex concepts into practical, real-world applications. The course struck a strong balance between theory and hands-on learning, particularly around agile thinking and product design workflows. Even post-course, Seann continues to support my product design journey through interview preparation, portfolio feedback, CV and cover letter reviews. I’d highly recommend this course & Seann to anyone looking to build or strengthen their skills in digital product design.”
“When I began working with Sai Constantine at IDNZ, I was studying Fashion Design (Honours) at Massey University and had only a very conceptual idea for a leather bag business. Through mentoring and brainstorming, Sai helped me expand my vision beyond clothing into accessories and supported me in developing both my creative and business skills, including idea iteration, market research, and branding. The programme provided valuable structure and guidance, introducing tools such as user personas, competitor analysis, and rapid design methods, while also helping me build a brand identity with clear values and direction. As a result, I developed a stronger brand story, clearer product range ideas, and thorough documentation of my design process, giving me a solid foundation to continue in the future. I would highly recommend Sai’s mentorship at IDNZ to other design students, as the programme can help transform uncertainty into a clear, tangible, and personal creative direction.”
Learning portal
Open 24/7, with everything included.
The IDNZ learning portal holds every module’s reading material and activities, and it’s where you hand in your work. It’s open 24/7 and all included in the fee, alongside unlimited one-to-one mentoring for the 8 weeks.
- Modules
- 8
- Portal
- Open 24/7
- Reading and activities
- All included
- Mentoring
- Unlimited, 8 weeks
- Reading for every module, written for working designers
- Hands-on activities for every module, practised in the live tool studio
- Your weekly hand-ins and your mentor’s feedback, in one place
- Updated every two months, with a dated note of what changed
- “Tool drop” sessions after major releases of Figma, Claude or ChatGPT, for 12 months
Price and intakes
The founding cohort starts 15 November 2026.
A new intake starts every two months after that. Paid tool access, unlimited mentoring and 24/7 learning portal access, with all reading material and activities, are included in every price below. All prices exclude GST unless shown.
Your price
- StandardAny intake
- $4,789 + GST$5,507.35 incl. GST
- Founding cohortThe 15 November 2026 intake only, 15 seats
- $3,997 + GST$4,596.55 incl. GST
- CAUXPDE graduatesStep-up price for graduates of the 11-week UX certification
- $2,397 + GST$2,756.55 incl. GST
- UX + AX pathwayCAUXPDE, then CAAXDE, booked together
- $7,186 + GST Save $2,392$8,263.90 incl. GST
- Other IDNZ certification graduatesCADME, CASME, CASMME or CADSE: $500 alumni credit
- $4,289 + GST$4,932.35 incl. GST
- One-day masterclass buyersYour masterclass fee comes off if you enrol within 3 months
- $489.99 off
Team rates
- Teams of 3 to 5From one employer: 10% off each seat
- $4,310.10 + GSTeach
- Teams of 6 or moreFrom one employer: 15% off each seat
- $4,070.65 + GSTeach
- Private cohort8 or more from one organisation, with your own product as the project
- Quote on request
The IDNZ UX course ladder
Start where you are. Step up when you’re ready.
The full UX process
CAUXPDE
11 weeks. $4,789 + GST, or $2,557 + GST after the JumpStart.
UX & product design certification
Already a practising designer? Start at CAAXDE. If you take CAUXPDE afterwards, CAAXDE counts towards it.
AX terms
The words you’ll use, in plain English.
AI design comes with new vocabulary. These are the terms CAAXDE uses, and what each one means in day-to-day design work.
- AX designAI and agent experience design
- Designing how people experience AI features, and how AI agents experience your product. It adds trust, disclosure and correction for AI features, and agent readiness, to the UX process.
- Agent experience
- How easily an AI agent acting for a person can find, read and finish tasks on a website or app, such as booking, paying or applying.
- MCPModel Context Protocol
- An open standard, introduced by Anthropic in 2024, that lets AI apps connect to other tools and data. Figma’s MCP server lets coding agents such as Claude Code and Codex read your design file, including its components and variables.
- Figma Make
- Figma’s AI tool that turns a prompt or an existing design into a working, clickable prototype.
- Claude Code and Codex
- AI coding agents from Anthropic and OpenAI that write and run code from plain-language instructions. In CAAXDE you use them to build prototypes from your Figma designs, then review every result.
- AI evals
- Tests that check an AI feature’s answers against agreed examples before people rely on it, and keep checking after launch.
- Golden set
- A set of example inputs with the answers you’d accept, used to score an AI feature the same way every time.
- Human in the loop
- A design where a person reviews, approves or can override what an AI does before it takes effect.
Questions
Questions about CAAXDE.
UX and AX, AI and design jobs, Figma’s MCP server, experience, hours, tools, coding, employers paying, assessment, NZQA and intakes.
Not sure CAAXDE is the right step?
Talk it through with a mentor. We’ll tell you honestly whether to start here, with CAUXPDE or with the JumpStart.
Book a consultationSee the 11-week CAUXPDE
What is AX design?
AX stands for AI and agent experience design. It covers designing AI features people can trust, designing products that AI agents can use, and building with AI agents. CAAXDE teaches all three on top of the research and testing UX designers already do.
What’s the difference between UX design and AX design?
UX design is about how people experience a product. AX design adds two more things to design for: the AI features inside the product, and the AI agents that use it on someone’s behalf. The research, testing and judgement from UX stay at the centre. CAAXDE adds trust patterns, agent readiness, AI evals and agent builds on top.
Does CAAXDE include design thinking?
Yes. Design thinking is built into CAAXDE, and your project runs through all five stages: empathise through five real interviews in week 2, define with an opportunity solution tree, ideate and prototype with flows, Figma Make and agent builds in weeks 3 to 5, and test with five users and AI evals in week 7. At each stage you learn where AI helps and where your own judgement has to lead.
Will AI replace UX designers?
AI is taking on more of the production work, such as first drafts of screens and working prototypes. Deciding what to build, talking to real people, judging whether an AI feature can be trusted and taking responsibility for the result still need designers. CAAXDE trains that side, and shows you how to hand the production work to AI safely.
What is Figma’s MCP server, and how is it used in CAAXDE?
MCP (Model Context Protocol) is an open standard that lets AI apps connect to other tools. Figma’s MCP server lets coding agents such as Claude Code and Codex read your design file, including its components and variables. In week 5 you connect your own Figma file this way and build a prototype, which you test with five users in week 7.
Do I need UX experience before CAAXDE?
Yes. CAAXDE is for people who already work in UX, UI or product design, or who have finished a UX course such as CAUXPDE. If you’re new to UX, start with the 5-week Agile UX Design JumpStart, then CAUXPDE.
How many hours a week does CAAXDE take?
About 10 hours a week for 8 weeks, around 80 hours in total. Each week you book live one-to-one sessions with your mentor, join the hands-on tool studio, and do the reading, activities and project work in the learning portal. Week 0, before week 1, is for setting up your tools, getting your project brief and meeting your mentor.
Do I need to pay for Figma, Claude or ChatGPT?
No. Paid access to Figma, Claude and ChatGPT is included for the 8 weeks of the course, so you can use every feature we teach. After the course you can keep using the free plans or buy your own.
How do the live sessions with my CAAXDE mentor work?
Your live lessons are one-to-one sessions with your mentor. You book them through the calendar we give you when you enrol: unlimited sessions Monday to Friday, with limited slots on Saturday and Sunday. You also have required check-ins in weeks 2, 4, 6 and 8.
Do I need to code to build prototypes with Claude Code or Codex?
No. You direct the agent from your Figma file through Figma’s MCP server and review what it builds. Week 0 sets up everything you need, and the week 5 studio takes you through your first agent build step by step.
Can I use my company’s product as the CAAXDE project?
Yes, with your employer’s agreement. Otherwise you work on a live brief from a New Zealand partner organisation. Partners are announced each intake.
Can my employer pay for CAAXDE?
Yes. We can invoice you or your employer: email godigital@id.ac.nz. Teams of 3 to 5 from one employer get 10% off each seat, and teams of 6 or more get 15% off. For 8 or more people, ask about a private cohort with your own product as the project.
How is CAAXDE assessed?
On your work. You hand in something every week, which builds your portfolio. In week 8 you present your final case study at the showcase and complete a timed practical. There’s no 200-question exam.
Does CAAXDE count towards CAUXPDE?
Yes. CAAXDE is an IDNZ industry certification, and if you take the 11-week CAUXPDE afterwards, CAAXDE counts towards it. Going the other way, CAUXPDE graduates can step up to CAAXDE for $2,397 + GST.
How does CAAXDE keep up with AI tools compared with NZQA-approved programmes?
CAAXDE is an IDNZ industry certification, so it’s reviewed and updated every two months, with a dated note of what changed. NZQA-approved programmes go through formal approval first: for a new degree programme, NZQA asks providers for at least three months’ notice and aims to complete approval within six months of the application. That process gives you an NZQA qualification and suits people who need one. CAAXDE is built to keep pace with the tools.
Is CAAXDE NZQA-accredited?
No. CAAXDE is an industry certification from IDNZ, not an NZQA qualification, and it isn’t eligible for Fees Free. It’s built for working designers who need practical skills and portfolio work they can show after 8 weeks of part-time study.
When are the CAAXDE intakes?
The founding cohort starts on 15 November 2026, and enrolments close on Wednesday 11 November. After that, a new intake starts every two months.
What is the founding cohort price for CAAXDE?
$3,997 + GST ($4,596.55 including GST) for the first intake on 15 November 2026, which is limited to 15 seats. The standard price is $4,789 + GST ($5,507.35 including GST).
Should I do an 8-week part-time AI UX course or a full-time UX bootcamp?
They suit different people. A full-time bootcamp usually suits someone changing careers who can stop working for several months. CAAXDE is for designers who already do UX and want to add AI and agent design while they keep working: about 10 hours a week for 8 weeks. If you’re new to UX and want to keep working, IDNZ’s JumpStart and CAUXPDE are part-time too.
What’s in the CAAXDE learning portal, and how often is the course updated?
The IDNZ learning portal is open 24/7 and holds each module’s lessons, reading material and activities, and it’s where you hand in your work. All reading material and activities are included in the course fee. The course is updated every two months, and each update comes with a dated note of what changed.
What’s the difference between CAUXPDE and CAAXDE?
CAUXPDE covers the full UX and product design process over 11 weeks, for people moving into UX or stepping up to product design. CAAXDE is 8 weeks for designers who already know that process and want the AI and agent layer: designing AI features and agent-ready products, and building prototypes with AI agents.
Does CAAXDE cover AI rules and Māori data sovereignty?
Yes. Week 8 covers New Zealand privacy rules, EU AI disclosure requirements and Māori data sovereignty, and how they shape the AI features you design. Week 6 covers trust, disclosure and correction, using Microsoft’s 18 guidelines for human–AI interaction.
Founding cohort
Starts .
$3,997 + GST for the first intake, with tools, mentoring and the learning portal included. Enrolments close Wednesday 11 November.

