Cover
Introduce AI to Azure to allow users' full creative freedom
Year
2023→2025
At
Microsoft Azure
Role
Product designer
Tool
Figma, Copilot
Location
Seattle, WA
Interactive prototypes
User flow
AI intergration
View product site
01
My team integrated AI to existing Azure experience to empower user build anything on their mind
AI can guide new users to learn about Azure
AI debugging
Create new workload with AI
02
Discovery of the Right Workload
Without guidance, new users don’t know which workload to start with, or still spending a long time looking.
Debugging Blind Spots
When something fails, users get cryptic error messages without clear next steps slowing down their progress.
Workload Fit & Tailoring
Even with workloads, advanced users sometimes can’t find an exact fit — they need more personalized options.
03
04
DESIGN
05
After the implementation of AI suggestions and AI debugging on Azure
Increase in traffic from AI suggestions to deployment
Users resolve errors with AI debugging
Abandoning their task after hitting a failure
06
Pre-flight validation
Auto-checks before run. If any fails, we don’t run. Every step is auditable.
Always review changes
AI generates a change set with reasons. Nothing runs until users accept per change or apply all.
De-risk uncertainty and hallucination
Low confidence or policy risk → auto-fallback to structured flow.
Easy rollback
One-click Undo.
