AI Builder
Helping makers create and use AI models without writing code. I played a foundational design role in AI Builder, from the product vision to the experience of configuring and using models.

Business users need ways to use their own data and expertise without first learning how to build a machine-learning system.
I partnered with PMs on the vision and led design work on model creation, usability and integration with the Power Platform.
Prebuilt and custom models support tasks such as document processing, object detection and prediction in apps and flows.
The product introduces AI through concrete business scenarios and guides makers toward a usable model.
The starting surface presents recognizable model types, so makers can choose the capability that fits their process.
For object detection, makers tag objects in their own images. The training task is expressed through a visual interaction.
A model detail view brings performance and the next available actions together before the model is used in an app or a flow.
The design connects familiar business tasks, understandable interactions and a shared product language.
I designed the no-code experience around model types and business examples. Form processing, for example, can be introduced through the document it will understand.
Working with PM, engineering and research, I aligned model experiences with the broader product direction and refined features through user feedback.
The goal extended beyond creating a model: makers needed to use its output in Power Apps and Power Automate to support their business process.
The project’s documented impact spans manual processing, workflow automation and access to reusable AI capabilities.
Intelligent automation reduced the manual data entry required in business processes.
AI Builder accelerated workflow automation across the industries described in the project.
The platform expanded access to capabilities including document processing, image detection and image classification.
