those who have a library with duplicates and hardcoded values and want to bring their tokens to a single standard
You will learn to run an AI-powered library audit, identify duplicates and out-of-sync tokens, and bring components to a functional, documented state.
those who work as a UI/UX or Product Designer and want to grow into a Design System Designer role
You will understand token architecture, the principles of component systems, and the basics of governance - and gain a concrete case for moving to a more systematic level of work.
those who lead a team and want to introduce AI into design processes without compromising quality
You will understand which design system tasks AI can take on - generation, auditing, and documentation - and get a plan for gradually introducing an AI-assisted approach within your team.
those who have to repeatedly explain colors and spacing in the absence of tokens
You will learn the 3-tier token model, learn to write structured prompts for Claude, and see how token names map to code through Figma variables.
the right course
- AI in design systems: where it helps (generation, transformation, documentation) and where design judgment is required
- shared glossary: LLM, prompt, context window, MCP, AI token vs design token
- token architecture: the 3-tier global / semantic / component model
- naming conventions and the W3C Design Tokens format
- modes as a separate dimension: light/dark and brand/region
- structured prompts for design system tasks: role, context, constraints, format, and example
- the difference between a vague and a structured prompt, with examples
- AI project context: the CLAUDE.md file
- linking Figma variables to code syntax
- practice: adding a new global palette to the Starbucks token system with Claude
- validating new tokens in Tokens Studio
- principles of component systems: atomic design in practice
- modularity, composition, variants, states, and responsive behavior
- library anti-patterns: detached instances, over-nesting, and hardcoded values
- duplicate components with different names: how to identify and avoid them
- generating a component specification with Claude: variants, states, props, and token usage
- building a Button component in Figma with all interactive states
- comparing AI outputs across participants - how prompt wording affects the result
- refactoring a component family: consolidating duplicates
- relinking detached instances and applying tokens
- generating TOKENS.md documentation with AI based on your own token system
- generating COMPONENT.md documentation: variants, states, token usage, and do/don't examples
- a practical audit framework: token consistency and component coverage
- checking drift, documentation accuracy, and naming consistency
- AI-assisted auditing: checklists, documentation gaps, and cross-referencing design and code
- auditing a real library with Claude and creating a fix list
- fixing tokens and validating them in Tokens Studio
- updating components: applying tokens and adding missing states
- updating TOKENS.md and COMPONENT.md to reflect the changes
- running a second audit pass to verify that the fixes worked
- governance basics: ownership and versioning of tokens and components
- keeping documentation in sync with the actual system
куратори
вони будуть поруч
A Portuguese product designer turned DesignOps enthusiast, currently focused on AI, design systems, and improving collaboration across teams.
Alongside her full-time job, she is an online teacher and actively supports the design community through Friends of Figma and Into Design Systems in Portugal. She is passionate about making design more effective within organizations by helping design, development and business teams work better together.
She is also interested in AI and how it can improve workflows and design systems, making them more efficient, scalable and impactful.
workshop price
350 EUR
pay by yourself
or through your employer
workshop price
350 EUR
pay by yourself
or through your employer
*According to the research of the Projector.
The study involved 567 respondents
among graduates of the courses in 2022.
got questions?
we have answers
- is this intensive for design beginners?
- do I need experience working with design systems?
- will AI do the design work for me?
- what tools are used during the intensive, and what do I need to have installed?
- why is Starbucks used as the basis?
apply
the first step is yours
Want to ask something else? Email to hello@prjctr.com or call +38 067 418-95-78.
Educational services are provided through access to public educational resources on the Internet, including virtual classrooms and an educational platform. The course falls under computer literacy training services within the meaning of clause 197.1.33 of the Tax Code of Ukraine.
the first step is yours