Mathias Davidsen: three AI workflows for product team prototyping — Miro

What the video covers

Mathias Davidsen, GM of Prototypes at Miro, walks through three AI workflows that change how product teams build, validate, and align on prototypes. The talk is grounded in what Miro’s tools can do as of May 2026 — including AI-generated prototype screens and Miro Sidekick, a voice-enabled AI assistant — but the workflow patterns described transfer to teams using other tools in similar ways. Each workflow is structured around concrete before-and-after comparisons.

Who it’s for

Product managers and designers who move through the prototype-to-validation cycle frequently and want to reduce the time between a rough idea and a tested, reviewable artifact. Particularly relevant for teams where the PM is responsible for coordinating feedback from both users and engineering before handing off to implementation, and where the limiting factor is how quickly prototypes can be prepared and reviewed.

Key takeaways

1. Prototyping without context-switching. The first workflow generates interactive screens directly within the same workspace where strategic context — research, decisions, problem framing — already lives. The practical benefit is that the prototype stays adjacent to the reasoning behind it rather than becoming detached as soon as the work moves to a separate design tool, which is a common failure mode when discovery and design happen in different places.

2. Synthetic usability testing at the design stage. The second workflow enables teams to run a form of usability testing on a prototype before any human participants are recruited. Davidsen describes going from an existing product screen to a synthetically tested and revised prototype without leaving Miro, which compresses the validation cycle and makes it easier to eliminate weak designs before spending time on actual user research sessions.

3. AI-assisted feedback synthesis for engineering handoff. In the third workflow, developers drag an asset from Cursor into the Miro canvas and team members leave feedback. Miro Sidekick then analyzes the collected comments — including subjective or contradictory notes — and generates a structured document with clear acceptance criteria. This addresses the manual work of reviewing feedback threads and translating qualitative input into specifications an engineering team can act on.

Worth watching if…

You frequently lose time between the end of a design review and the point when engineering has clear requirements, or if your prototypes are often disconnected from the user research that motivated them. The third workflow is particularly relevant for teams where cross-functional feedback from multiple stakeholders needs to be translated into something actionable before development can begin.