Top 10 Crypto Exchange · Neo-Banking · AI Design Ops · 2025

AI-Enhanced
Design
Operations

Built a four-layer AI workflow, Claude, Cursor, Figma MCP, that reduced design-to-production handoff by 70%, shipped 40+ micro-interactions that would never have made the backlog, and maintained 98% design system consistency while launching a neobank MVP to $7M+ in transactions.

Role

Head of Design

Timeline

Sep 2025 – Present

AI Stack

Claude · Cursor · Figma MCP

Early Impact

$7M+ transactions in 8 weeks

-70%
Design-to-production handoff time
98%
Design system consistency
40+
Micro-interactions shipped via AI
Operational Impact, Before & After AI Integration
DURATION8 weeks
SCOPEFull Design Ops Transformation
AI TOOLSClaude, Cursor, Figma MCP
Before
72%
Design System
Compliance
3 days
Handoff
Cycle Time
12x/wk
Design
Review Meetings
~60%
Spec Writing
Accuracy
38 hrs
Weekly Overhead
per Designer
4x
Dev Spec
Revisions
Manual
Governance &
QA Process
Process: Design in Figma → Manual Spec Writing → Slack/Email → Dev Interpretation → Review Loops → Revisions Cycle
Live
After (AI-Integrated)
98%
Design System
Compliance
+36% increase
Same-day
Handoff
Cycle Time
-70% faster
7x/wk
Design
Review Meetings
-40% reduction
~98%
Spec Writing
Accuracy
+63% improvement
12 hrs
Weekly Overhead
per Designer
-68% reduction
<1x
Dev Spec
Revisions
-75% fewer
Automated
Governance &
Real-time QA
Proactive
Process: Design in Figma → Figma MCP Auto-Export → Claude AI Spec Gen → Cursor Integration → Auto-QA Check → Ready-to-build
AI Stack & Infrastructure
Claude
• Spec generation
• Interaction flows
• Copy refinement
• Jira automation
Cursor IDE
• Real-time code assist
• Component generation
• DevTools integration
• File sync automation
Figma MCP
• Token export automation
• Component API access
• Design-to-code bridge
• Live sync with Cursor
Transformation Timeline
Week 1-2
Audit Current State
Pain point mapping
Bottleneck analysis
Week 3-4
Implementation
AI workflow setup
Tool integration
Week 5-6
Pilot & Refinement
Team training
Workflow optimization
Week 7-8
Optimisation & Scale
MVP launch
Governance setup
Team Capacity Impact
Product Designer
18 hrs/week
Reclaimed from spec writing & rework
Front-end Eng
12 hrs/week
Fewer spec clarifications & rework
Design Lead
8 hrs/week
Fewer review meetings & escalations
Product Manager
6 hrs/week
Automated status & comms synthesis
Total Time Reclaimed: 44 hours/week across team (equivalent to 1.1 FTE)

Four-layer
AI
workflow.

I architected an AI-augmented design workflow where multiple AI systems worked together to eliminate friction at every stage of the design-to-delivery pipeline.

01

Layer 1: Figma MCP + Cursor, Design System as Code, Connected Cursor to our Figma design system via MCP, creating a live bridge between tokens, components, and code. Design-to-code handoff time reduced by 70%. Zero interpretation gaps between design and engineering.

02

Layer 2: Claude Code, Interaction and Animation Refinement, Used Claude to prototype and perfect micro-interactions in code: card activation sequences, balance transitions, navigation choreography. Shipped 40+ refinements that would have been deprioritised in traditional workflows.

03

Layer 3: Claude + Jira/Slack, Operational Automation, Configured AI integrations to auto-generate status updates from Jira, draft stakeholder comms, and synthesise design review notes. Reclaimed 8+ hours per week. Reduced meeting cadence by 40%.

04

Layer 4: AI-Monitored Design System Governance, Cursor, via Figma MCP, flags component deviations in real-time before code review. Proactive quality gate, not reactive cleanup. Design system adoption: 72% → 98%.

05

MVP launch execution, All four layers operating in parallel while shipping the full neobank: onboarding, payments architecture, debit card lifecycle. $7M+ in transactions within 8 weeks.

"Every tool in this stack, Claude, Cursor, Figma MCP, is available to anyone. The competitive advantage isn't the tools. It's knowing where to point them."

Craft
at
velocity.

The AI-augmented workflow proved that pace and craft aren't trade-offs, they're multipliers when AI eliminates the friction between intent and execution.

Handoff Speed

70% reduction in design-to-production handoff time across the entire product.

Craft at Scale

40+ refined micro-interactions shipped, details that delight customers and would never have made a traditional backlog.

System Consistency

98% design system adoption (up from 72%) with zero regressions shipped to production.

Launch Impact

$7M+ in transactions within 8 weeks of MVP launch. 30K+ debit cards ordered.

Next Case Study

Metro Bank:
The UK's First AI in Retail Banking

Want in?

Score 100% on the Design Eye test to earn access. Or enter the password if you have it.

4 rounds. No partial credit.