What we’re learning about AI in our design work — the tools, workflows, experiments, and lessons from across the Automattic Design team.
80+
Designers
24
Countries
4
AI tools in rotation
1
Team learning in public
The Toolkit
What we reach for
Figma AI / Make
First-pass layout generation and visual exploration. Strongest for rapid wireframing and variant generation. Still learning where it hits a ceiling for production-quality output.
Claude
Prototyping, UX copy iteration, and design documentation. Excels at turning rough thinking into structured output. Used daily for everything from copy variants to code scaffolds.
Midjourney
Moodboarding, concept art, and visual direction. Fast path from research to visual language. Best when you know the aesthetic territory you’re heading into.
GitHub Copilot
Design-to-code workflows and prototype implementation. Bridging the gap between mockup and working code — particularly for WordPress block development.
Workflows & Recipes
How we’re working
Step-by-step processes contributed by designers across the team. Practical, repeatable, and honest about what works and what doesn’t.
NO.1
Research to moodboard
How to go from a research dump to a moodboard in an afternoon using Midjourney. Start with keywords, refine with style references, and curate with intention.
NO.2
UX copy variations
The Claude prompt that actually works for generating UX copy variations. Structured input, tone constraints, and how to iterate without losing voice.
NO.3
Design to code
Using Copilot to turn a Figma mockup into a working WordPress block prototype. Where it saves time and where you still need to intervene.
NO.4
Your workflow here
Got a process worth sharing? We’re collecting contributions from across the design team. DM or comment below to add yours.
Experiments
Learning in public
-
Telex experiments
“Describe your idea. Telex will build a WordPress block for you.” — That’s the promise featured on the Telex website. Created by Automattic, this experimental tool has been generating buzz across the WordPress community as a powerful new companion for developers.
-
Content for fun vs. content for purpose: designing for two distinct modes of consumption
From AI assistants to digital platforms, how can we design for rapid mode switching in real life? Reflections about utilitarian and experiential content and why understanding both matters.
-
Behind the scenes: How we iterate on UI copy
A behind-the-scenes look at how Automattic designers treat UI copy as a core design element—from first drafts to faster iteration with internal tools like Wordmattic. Learn how writing early, often, and in context shapes clearer, more human interfaces.
-
From idea to innovation: write brief with AI
Read how a side project turned into a product feature, and launched in just two months.
AI IS MOVING FAST. WE’RE LEARNING TOGETHER AND SHARING WHAT WE FIND.
— Automattic Design
Learning together
Building AI fluency
Where we’re investing in more structured ways to build AI fluency across the design team. Connected to two programmes running this year: the AI-centred Design Meetup and the AI Enablement Programme in New York.
NO.1
Design Meetup
An in-person gathering where designers share what they’ve been building with AI, swap techniques, and push each other’s work forward.
NO.2
AI Enablement
A structured training programme running in New York, May 2026. Hands-on sessions, shared prompts, and practical frameworks.
NO.3
Field Notes
This page. A living, crowd-sourced resource that grows as we learn. Every contribution makes it more useful for the next person.
NO.4
What’s next
Potential integration with Automattic.ai, cross-posting from the AI Enablement Programme, and a lightweight editorial process.
Contribute
Share your field notes
If you’ve run an AI experiment, built a workflow worth sharing, or have a tool recommendation — we want to hear from you. This page is crowd-sourced by design.
Automattic Design 