Natera

optiFLOW: Validating AI-Generated Document Tags

optiFLOW: Validating AI-Generated Document Tags

As a UX Designer partnering with the Senior Product Lead for Billing Experience, I designed a clearer workflow for reviewing AI-generated document tags, supported by presentation assets, system diagrams, and final UI.

Explore the prototype below by adding or removing a tag. The design system documentation is embedded in this recreated prototype.

My Role

Product Designer

Team

UX Designer (me), Senior Designer, Engineer, PM

Stack

Figma, Claude, Gemini, Miro, UX Thinking, Figma Make

Company

Natera

Industry

Biotechnology

Timeline

2 weeks

My Role

Product Designer

Team

UX Designer (me), Senior Designer, Engineer, PM

Stack

Figma, Claude, Gemini, Miro, UX Thinking, Figma Make

Company

Natera

Industry

Biotechnology

Timeline

2 weeks

Introduction

Introduction

The existing review process relied on fragmented manual steps, making AI-generated tags slower to validate and harder to correct. Working with the Senior Product Lead for Billing Experience, I explored a clearer review screen and extended the platform with new tools, components, and interaction patterns.

  • New tools: Controls for accepting, removing, correcting, and validating tags

  • New documentation: Component behavior, variables, states, and edge cases

  • Team alignment: A shared reference for product, design, and engineering

If implemented, the proposed workflow could reduce review time and rework while improving tag consistency and document throughput. These outcomes are hypothetical and would need to be validated through usability testing, workflow analytics, and production data.

read more about my process, design documentation, and designing around assumptions

The existing review process relied on fragmented manual steps, making AI-generated tags slower to validate and harder to correct. Working with the Senior Product Lead for Billing Experience, I explored a clearer review screen and extended the platform with new tools, components, and interaction patterns.

  • New tools: Controls for accepting, removing, correcting, and validating tags

  • New documentation: Component behavior, variables, states, and edge cases

  • Team alignment: A shared reference for product, design, and engineering

If implemented, the proposed workflow could reduce review time and rework while improving tag consistency and document throughput. These outcomes are hypothetical and would need to be validated through usability testing, workflow analytics, and production data.

view case study deck here

The existing review process relied on fragmented manual steps, making AI-generated tags slower to validate and harder to correct. Working with the Senior Product Lead for Billing Experience, I explored a clearer review screen and extended the platform with new tools, components, and interaction patterns.

  • New tools: Controls for accepting, removing, correcting, and validating tags

  • New documentation: Component behavior, variables, states, and edge cases

  • Team alignment: A shared reference for product, design, and engineering

If implemented, the proposed workflow could reduce review time and rework while improving tag consistency and document throughput. These outcomes are hypothetical and would need to be validated through usability testing, workflow analytics, and production data.

read more about my process

Copyright © 2024

Copyright © 2024