Designing Trust Into a $1M Decision
Led Human-AI decision design for Fortune 500 Tech—43% faster time-on-task, 144× faster data access, zero blind automation on $100K–$1M+ decisions.
*Content has been de-identified in accordance with the NDA.
Industry
B2B
SaaS
Enterprise UX
AI/ML
Human-AI Interaction
Internal Tools
Data
API
Role
Senior Product Designer
Team
2 Product Designers • 2 UX Researchers • 1 PM • 5 Developers • 3 Data Scientists
Skills
Wireframing • Prototyping • Interaction • Scalable Design Systems • Stakeholder/User Interviews • Personas • Journey Mapping • User Flows • Usability Testing • A/B Testing • Research Synthesis • Accessibility (WCAG) • AI-Assisted Design-to-Code Workflow
Impacts
01· The Challenge
The alert was buried. The AI recommendation wasn't trusted.
"How might we help an engineer notice, trust, and act on the alert that matters—before it becomes a $1M problem—without asking the user to hand judgment to a system they've never seen be wrong?"
02· The Research: Where Trust Already Breaks
16 in-depth user interviews synthesized into persona, journey map, and insights.
Frictions
→ Opportunities
Alert sits three levels deep in the tree, and the AI Update read like a plain column label.
Surface the alert and the recommendation above the table—nothing to dig for.
No visible reasoning—reads as a black box.
Show the pattern match and the incident history behind it, not just a score.
Confirm looks obvious—so people stop checking, even when the AI could be wrong.
The user weighs the evidence first—Override stays as easy as Confirm, and a thumbs vote helps it improve next time.
03· The AI-Powered Workflow: Human-AI Interaction
The AI never acts alone. The user weighs the evidence before deciding—Override stays just as easy as Confirm.
04· Test and Validation: Rounds of Usability Testing, A/B Testing on AI Recommendation, Banner Cards
05 · The Solution: Shipped AI-Powered Design and Scalable Design Systems
I’ve built reusable components and tokens for scalable design systems that AI agents can read and deploy, enabling a seamless, AI-assisted design-to-code workflow in complex, data-heavy industries.
Implement scalable capabilities using a systems-thinking approach, collaborating with developers and data scientists to deliver B2B SaaS solutions for technical professionals in enterprise UX to support $9-figure industrial operations.
Full case studies on the AI recommendation and AI assistant: how I raised CSAT from 53% to 92%, the technical constraints and trade-offs behind each design. Let’s chat live!
Impacts