Year

2025

Timeline

16 weeks

Azure OpenAI Services App Creation Flow

Helping developers get started with AI by doing, not just reading.

The Azure Portal already provides developers with a full toolbox for building cloud solutions. But for first-time AI builders, getting started could still feel daunting. In early 2025, I led design for the new App Creation Flow, a pre-packaged project kit that layers on top of the existing portal experience. By bundling recommended services with a one-click sample app, we reduced setup time from hours to minutes, creating an experience that was up to 95 percent faster. The flow was later adopted as a reference pattern across the Azure Portal.

This project was not about a flashy visual identity. It was about solving a difficult information architecture and workflow problem. The Azure Portal already gave developers every tool imaginable, but if you did not have an app to deploy, there was no clear next step. Our challenge was to organize configurations, service options, and deployment paths into something approachable and obvious, especially for first-time AI builders.

My Role

  • Lead Product Designer – defined the interaction model, visual hierarchy, and component styling. I also collaborated with engineering, research, and PMs across three time zones, translating complex constraints into clear design opportunities.


Design Goals

  • Provide a clear and approachable next step for users who do not yet have an app to deploy

  • Help developers learn through hands-on experience rather than reading documentation alone

  • Streamline setup while still allowing flexible configuration for advanced users

Iteration
As we refined the App Creation Flow, we revisited how the Recommended Services feature appeared in the portal. While the capability already existed, its earlier version was static and easy for users to miss. We redesigned the experience so that recommendations dynamically reflected project context, surfacing relevant services such as Key Vault, App Configuration, and SQL Database at the right time during setup.


This update gave developers better guidance and a clearer sense of what each service contributed to their deployment. It connected the AI sample app work to the broader Azure experience, making the overall setup process more contextual and intelligent.

Impact

  • Reduced AKS app deployment time from hours to minutes, improving speed by roughly 80 to 95 percent

  • Shipped to the Azure Portal in 2025 and now used by thousands of developers who are getting started with Azure OpenAI

  • Refined how Recommended Services are surfaced, introducing logic that reads the user’s Azure subscription to present smarter, more relevant service options

  • Strengthened the foundation for future guided experiences across the Azure Portal by establishing a reference pattern for workflow-first design

Documentation

Learn more about the AOAI sample app


Designed in Chicago
Serving customers worldwide
Designed in Chicago
Serving customers worldwide
Designed in Chicago
Serving customers worldwide