# How a UX researcher uses Figma Make at GM Source: https://www.youtube.com/watch?v=gU8FqJDxO_0 Recap page: https://rapidrecap.app/video/gU8FqJDxO_0 Generated: 2026-01-28T13:32:58.783+00:00 --- ## Quick Overview UX Researcher Laura Dunn from General Motors demonstrates how the Figma Make AI tool significantly streamlines the creation of complex design artifacts like detailed user flows and design system documentation by generating high-fidelity outputs directly from natural language prompts, saving substantial time compared to manual iteration. **Key Points:** - Laura Dunn, UX Researcher at GM, showcases using Figma Make AI to generate a comprehensive vehicle service scheduling flow based on screenshots and specific requirements. - The AI successfully generated a multi-step, single-page flow incorporating progressive disclosure, which was a major goal of the initial prompt. - The initial prompt was detailed, outlining 5 steps for the service flow, including selecting services, location, dealer/address, date/time, and contact confirmation. - The AI-generated output maintained the Buick brand aesthetic, including color palettes and typography, demonstrating adherence to brand guidelines. - Laura noted that while the AI output was impressive, it sometimes required manual iteration, such as refining the navigation pattern documentation that was generated separately. - The AI helped rapidly prototype complex interaction patterns like conditional visibility based on selected services, which would have taken significantly longer manually. - The session concluded with Laura demonstrating how the AI helped refine the navigation pattern guidelines documentation by producing a highly structured, multi-column table based on complex input. ![Screenshot at 00:03: The screen shows Laura and Jay discussing the initial goal: using Figma Make to illustrate the overall architecture of the Buick service flow based on existing design screenshots.](https://ss.rapidrecap.app/screens/gU8FqJDxO_0/00-00-03.jpg) **Context:** The video features a discussion between UX Researcher Laura Dunn from General Motors (GM) and host Jay Dalal on the "Deep Dive" channel, focusing on practical applications of the Figma Make AI tool. Laura uses screenshots of an existing Buick service scheduling website flow to prompt Figma Make to create an improved, interactive, single-page scheduling experience that adheres to GM/Buick branding and incorporates progressive disclosure principles. ## Detailed Analysis The video details a practical demonstration of using Figma Make AI, specifically by GM UX Researcher Laura Dunn, to overhaul a complex vehicle service scheduling flow from Buick.com. Laura provided the AI with existing design screenshots and a detailed, numbered prompt outlining the desired five-step, single-page flow that emphasizes progressive disclosure. The AI successfully generated a substantial portion of the required flow, including the service selection step with conditional visibility (e.g., showing 'At Home Service' availability only for certain services). Laura highlighted that the AI output was impressive in adhering to the Buick brand aesthetic and fulfilling the core requirement of a single-page flow, saving significant time compared to manual wireframing. The discussion also touched upon using the AI to generate supporting documentation, such as a comprehensive table comparing horizontal, vertical, and hybrid navigation patterns, which the AI populated based on criteria like IA depth, user goal, and scalability. Laura noted that while the AI was highly effective, especially for rapid prototyping, some outputs required manual refinement, such as the navigation guideline table, suggesting that while the tool is powerful, human oversight remains crucial for pixel-perfect execution. ### Service Scheduling Flow Generation - AI generated a multi-step, single-page flow using screenshots as context - AI correctly implemented progressive disclosure based on service selection - AI adhered to Buick's brand aesthetic in the generated layout ### AI Output Quality & Iteration - AI output was highly impressive and prescriptive, saving significant time - Required some manual iteration, particularly on complex documentation like navigation guides - User feedback loop demonstrated rapid iteration capability ### Navigation Pattern Guidelines Generation - AI created a detailed three-column comparison table for Horizontal, Vertical, and Hybrid navigation patterns - The table successfully incorporated complex criteria (IA depth, user goal, scalability) and benefits/limitations - Laura felt the AI output for this documentation was better than manual creation. ### Key Takeaways - AI is excellent for rapid prototyping of complex flows and generating supporting documentation - Precise prompting is key to achieving desired fidelity and adherence to constraints - The process is highly collaborative, allowing designers to iterate quickly on AI-generated assets. ![Screenshot at 00:04: The initial Figma canvas showing the original Buick service flow screenshots being analyzed by Laura and Jay.](https://ss.rapidrecap.app/screens/gU8FqJDxO_0/00-00-04.jpg) ![Screenshot at 00:34: The AI-generated single-page 'Schedule Service' interface showing the service selection step with checkboxes and 'At Home available' indicators.](https://ss.rapidrecap.app/screens/gU8FqJDxO_0/00-00-34.jpg) ![Screenshot at 01:12: The AI generating the 'Select Vehicle' screen as part of the flow reconstruction process.](https://ss.rapidrecap.app/screens/gU8FqJDxO_0/00-01-12.jpg) ![Screenshot at 03:08: Laura pasting the detailed prompt into the Figma Make AI command bar to initiate the design generation.](https://ss.rapidrecap.app/screens/gU8FqJDxO_0/00-03-08.jpg) ![Screenshot at 03:56: The AI generating the Navigation Pattern Guidelines table on a separate canvas, demonstrating its ability to create documentation.](https://ss.rapidrecap.app/screens/gU8FqJDxO_0/00-03-56.jpg)