3 AI Workflows Changing How Product Teams Prototype

In this video, Mathias Davidsen, GM of Prototypes at Miro, walks through three AI workflows changing how products get built, validated, and shipped. Whether you're a product manager, designer, or engineer, this is how AI is starting to replace weeks of prototyping work in a single afternoon, without breaking your design system or losing user context. What's the worst usability bug an AI has caught in something you shipped? Drop it in the comments, we read every one. 🔥 What you'll see: → How to evolve existing products with AI by pulling PRDs from Confluence directly into your Miro board → How Miro Sidekick runs a forensic usability audit against Nielsen's 10 Heuristics and surfaces 33 violations in a single prompt → How to hand a Miro prototype to Claude Code via MCP and get a pixel-perfect React app on the other side ⏱ Chapters: 0:00 — The three AI workflows changing prototyping 0:32 — Workflow 1: Pulling a PRD from Confluence into Miro 1:46 — Implementing the PRD into the prototype with Sidekick 2:53 — Iterating with AI: from sidebar to horizontal layout 3:56 — Workflow 2: Synthetic validation with Nielsen's 10 Heuristics 4:55 — The audit surfaces 33 distinct violations 5:32 — Generating a new prototype from the AI's findings 6:48 — Workflow 3: Handoff to build (Figma or Claude Code via MCP) 8:00 — Claude Code generates a working React app 8:22 — What's next + Canvas 26 event invitation 🛠 Tools featured: - Miro Sidekick — Miro's native AI in the canvas - Claude Code (via Mrio MCP) - Confluence (PRD source) - Miro MCP 🟡 About Miro: Miro is the AI innovation workspace. Over 100 million people use Miro to imagine, plan, and build the future of their products and businesses. ▶ Subscribe for more on the future of product, design, and AI: https://youtube.com/@MiroHQ #AIPrototyping #MiroAI #AIDesign #ProductDesign #AIWorkflow