What an agent needs from your design system in Markdown
A structured text version of the portfolio to read, share or use as context in an AI assistant.
_Services_
# How I can help
Three ways I work with product, design and development teams. Each one starts with a real problem from your team and ends with something you can keep using.
_01 / UX_
## Product design and UX
Make a complex product clear to the people who use it.
Related work
- Complete redesign of My Sunrise (https://baguelus.com/en/proyectos/my-sunrise/)
- Sunrise Rewards (https://baguelus.com/en/proyectos/sunrise-rewards/)
Talk about this (https://baguelus.com/en/contacto/)
### It fits when
- A portal or internal tool has grown feature by feature and people get lost in it.
- You are about to redesign and want evidence before committing budget.
- Business, technology and users are pulling in different directions.
### What we do
- **Frame**: Agree which problem the product needs to solve and which decisions are open.
- **Understand**: Interviews and observation to see how people work and what constrains the team.
- **Decide**: Workshops that turn findings into priorities and things worth testing.
- **Prototype and validate**: Flows and prototypes tested with users, iterated against business and technical limits.
### What you get
- Research findings the team can act on
- Tested flows and prototypes
- A UI kit connected to your design system
- Decisions documented, with the points to check before development
- Accessibility considered from the start
_02 / DesignOps_
## DesignOps and design systems
One set of rules for designers, developers and AI agents.
Related work
- A design system for AI (https://baguelus.com/en/proyectos/design-system-ia/)
- What an agent needs from your design system (https://baguelus.com/en/servicios/)
Talk about this (https://baguelus.com/en/contacto/)
### It fits when
- Design and code drift apart after every release.
- Handover means explaining the same decisions again and again.
- You want AI agents to build with your components, not around them.
### What we do
- **Follow a real flow**: From Figma to code, to find where information gets lost or work is repeated.
- **Review the system**: Tokens, components, documentation and how changes are validated.
- **Write the contracts**: Component contracts and handover rules that people and agents can both read.
- **Test with your team**: Run the workflow on a real case and adjust before extending it.
### What you get
- A map of where the handover breaks today
- Tokens and component contracts
- Documentation readable by designers, developers and agents
- A tested workflow from Figma to React and Storybook
- A plan to extend it component by component
_03 / AI_
## AI adoption and training
Bring AI into the team’s daily work, one useful pilot at a time.
Related work
- Voice of Customer pilot at Sunrise (https://baguelus.com/en/sobre-mi/#ia)
- How to choose your team’s first AI pilot (https://baguelus.com/en/articulos/primer-piloto-ia/)
Talk about this (https://baguelus.com/en/contacto/)
### It fits when
- Everyone is trying tools, but nothing sticks in the workflow.
- You need to show what AI is worth before investing more.
- The team wants to learn with its own work, not with generic demos.
### What we do
- **Choose one task**: A bounded task with clear input and output, compared with how it is done today.
- **Define the criteria**: Agree what the team must review and how we will know the pilot is worth extending.
- **Build the pilot**: Prototypes, agents and automation, tested on representative examples.
- **Train the team**: Hands-on workshops where the team evaluates results and adjusts the workflow.
### What you get
- A working pilot on a real task
- Evaluation criteria and documented results
- Workshops built on your own examples
- A recommendation on tools and next steps
- Mentoring while the team takes it over
## Ways to work together
Most collaborations start small and grow from what we learn.
-
_01_
### Diagnostic
A short review of one product, flow or design system, with clear recommendations and priorities.
-
_02_
### Project or pilot
I lead or join the work from discovery to delivery, alongside your team.
-
_03_
### Workshop
Hands-on sessions built on your team’s own material, in person or remote.
-
_04_
### Mentoring
Regular sessions for designers, product people or leads who are changing how they work.
Based in Bad Wörishofen, near Munich. On site across the Zurich–Munich corridor and remote anywhere.
## Common questions
**Do you only advise, or do you also build?**
Both. I build the prototypes, design-system pieces and pilots myself, and train the team so the work continues without me.
**Can you work with our existing design system and tools?**
Yes. I start from what you already have and connect to it rather than replace it. I usually work with Figma, GitHub, React and Storybook.
**We are new to AI. Where do we start?**
With one small, repeated task and a clear way to judge the result. A first pilot shows more than a strategy document.
**How do we begin?**
Tell me what you are working on. A first conversation is enough to see which of these formats fits, if any.