Show HN: AI agents that run real user interviews

The idea is basically:

AI can build products now, but it usually relies on assumptions or synthetic feedback.

This lets an AI agent create a real user interview study, share a link with participants, and retrieve themes + verbatim quotes once interviews complete.

Works with Claude Desktop and Cursor via MCP.

Agents create a study, return an interview link to share with users, and retrieve themes + quotes.

Supports visual stimulus (images or Figma prototypes) for concept testing.