Can AI read a user flow drawn on a whiteboard?

User flows get sketched on whiteboards constantly in product and design work: during kickoffs, design critiques, product reviews, and sprint planning sessions. The board shows screens as boxes, decisions as diamonds, and transitions as arrows — often annotated with trigger events, edge cases, or error states noted in the margins.

This whiteboard artifact is valuable, but it's fragile. The team leaves, the board gets erased, and the shared understanding of the flow exists only in the designer's head.

What BoardSnap reads from a whiteboard user flow:

• Screen or state labels ("Login screen," "Dashboard," "Checkout")
• Transition arrows and any trigger labels ("Submit," "Back," "Skip")
• Decision points and their branch labels ("Logged in?" → Yes/No)
• Error state paths and recovery routes
• Annotations about behavior, timing, or backend dependencies
• Open questions flagged during the session

The output: BoardSnap returns a written description of the flow — step by step, including decision branches. This seeds a formal user flow doc in Figma, Miro, or Confluence without starting from a blank page. It also gives the developer a readable spec for what the flow should do before the polished Figma version exists.

Pro workflow: After snapping, open the project AI chat. Ask "What edge cases are missing from this flow?" or "How would this flow change for a returning vs. new user?" — the AI uses the board summary as its input.

Limitations: Flows with many overlapping arrows or very dense screen grids produce less precise output. For complex flows, snap the happy path first, then snap the error and edge-case sections separately.

For broader use case context, see the design review and product manager pages.