I wire AI into design teams. Not decks. Working setups.
Tools configured. Designers trained. Workflows handed off. Claude Code + Figma wired into design orgs. Now I do it for teams like yours.
// Built design systems for
Five problems every design team hits with AI.
None of these get fixed by another AI tool.
Design is where the team waits.
Three boring tasks eat your design capacity: research synthesis, documentation, asset production. Each one cuts from days to minutes when AI is wired into the right handoff point. The flashy demos don't touch them.
80 days1h
640×
faster design system audit
Competitors use AI. Your team doesn't.
One designer on your team already tests AI alone. The rest don't know where to start. Workshops get everyone to the same starting line in 1 to 2 days.
Everyone talks AI. Nobody shows the workflow.
Every audit answers three ways: implement this, skip that, AI doesn't help here. Most consultants won't say that last one.
Your team is growing. Your process isn't scaling.
One source of truth. Shared prompts. Zero guesswork. New hires onboard in days, not weeks, because the workflow isn't stuck in one person's head.
Hardcoded values ship.
An audit flags every hardcoded color, off-scale spacing, and broken component in 1 hour. With file references.
Rolling out Claude Code across a 50-person product team.
Design couldn't keep up with product.
A 50-person team building a full consumer product. Design system had ~1,200 components. Onboarding new designers took two weeks. Audit of consistency drift across the system took 80 days of senior time. Nobody wanted to run it.
- Slow handoff from design to engineering
- Token drift going unnoticed for weeks
- One AI enthusiast, 49 people waiting
Wired AI into the real process. Not on top of it.
Audit first, pitch second. Pipeline configured to their real design system, not a demo. Workshops ran on their own backlog.
- Figma MCP reading their component library
- Shared prompt library committed to team repo
- Token pipeline: Figma → code in one command
- Standardized handoff template for every project
Team shipped with AI the Monday after.
Everyone walked out with a configured setup. No "cool workshop, what now?" They were already using it on their backlog. Three months later the workflow is still live and new hires onboard in days.
week 1–2
Audit + workflow mapping
week 3–5
MCP pipeline + workshops
week 6–8
Rollout + knowledge transfer
Show, don't tell.
This is what I actually run on client projects. No magic. Just tools wired into your process the right way.
✓ Manual audit time: ~80 days of senior time✓ This workflow: ~1 hour end-to-end✓ Your team learns to run this themselves in 1 workshop session
Start small. Scale when it works.
Every engagement leaves a setup your team already shipped with.
Discovery
~ half day · on-site or remote
Half-day session, live on your actual design system. Walk out with a prioritized list of where AI helps, where it doesn't, and what to wire up first.
- Process map with bottlenecks
- 3–5 prioritized AI use cases
- Working prototype on your files
- 3-month roadmap with owners
→ Know what to implement
request_discovery →Workshops
~ 1–2 days · your office or remote
Your whole team, 1-2 days, on your own backlog. They leave with Claude Code + Figma MCP configured for your stack, shared prompts committed to a repo, and a workflow each designer used during the session.
- Everything from Discovery
- Working workflow per designer
- Shared prompts & templates
- Real tasks. No sandboxes.
→ Ship with AI Monday
request_hands_on →Support
~ monthly · async + check-ins
Ongoing support as your AI workflow evolves. Monthly check-ins, async support in Slack/email, new-hire onboarding to your workflow, and prompt updates when Claude Code + Figma ship major changes.
- Async support (Slack/email)
- Monthly check-in calls
- New hire onboarding
- Workflow iteration
→ Scale with your team
request_support →How this actually goes.
Every phase ships something concrete. You see it, you use it, you own it.
Discovery
A call + async review. Your tools, process, pain points on the table. Focus: where AI actually cuts shipping time.
call + async reviewAudit
Your workflow mapped. Priorities set: implement, skip, or leave alone. You walk out with an executable roadmap.
3-month roadmapBuild
Tools configured. Workflows built. Templates committed to your repo. Team trained on real tickets from your backlog.
working setupSupport
Monthly check-ins. Async Slack/email support. New-hire onboarding. Prompt updates when Claude Code + Figma ship major changes.
monthly check-in// but what does that actually change?
A typical cycle in design.
- 2 days
Audit design system for drift
- 2 weeks
Onboard a new designer
- 4 hours
Write handoff spec for one flow
- 1 day
Generate component variants
- 3 hours
Synthesize research interviews
The same cycle, but shipping.
- 3 min
Audit + drafted fix plan
- 2 days
Onboard with pre-configured setup
- 20 min
Spec drafted from Figma
- 40 sec
All variants + naming convention
- 10 min
Insights with cited quotes
Figma talks to Claude Code through MCP. Output lands wherever you want it.
// origin story
From fintech designer to AI wiring specialist.
For six years I built fintech products: BNP Paribas, Santander, VeloBank. Every project ended the same way. Solid design system. Smooth handoff. Worn-out team. Something was always slower than it needed to be.
Then in 2025, during a Claude Code + Figma MCP rollout with a fintech product team, I watched a 50-person team stop dreading design system audits. The workflow I wired up turned a months-long marathon into a one-hour task.
That's when I realized: design teams don't need another AI strategy deck. They need someone who has done the wiring, in production, with real people who push back. That's the job I took on.
What it's like to work with me.
In their words, not mine.
Dominik doesn't just focus on visuals. He thinks through the product and user experience end-to-end. Strong imagination, solid understanding of the market. Really great person to have on the team.
Krystian W.
// product manager
Dominik brings clarity to complex integrations and balances business and user needs effectively. A reliable partner in fast-moving fintech. I highly recommend him to any team working in fintech or integration-heavy products.
Sofia P.
// head of product
Your name here — after we ship something together.
[your name here]
// coming soon
Questions, answered.
What's slowing your design team down?
Tell me about your team. I reply within 24 hours.
Dominik Księżyk
// ai in design process
6 years designing fintech products: BNP Paribas, Santander, VeloBank. I configure tools, train teams, and leave behind working workflows. No decks.
// or email directly
// based in Poland
// working with teams worldwide
// currently booking July 2026