Telegram UI Design Workflow: Google Stitch + Gemini AI

Chat to generate Telegram-ready UI designs—using Google Stitch + Gemini AI

This n8n workflow turns natural-language prompts into UI design and prototypes through an AI-powered Telegram experience. By connecting Telegram, Google Gemini, and Google Stitch (MCP), it helps you create and manage design projects, list screens, and generate new UI screens—without manually wiring API calls.

What this workflow does

• Creates new design projects from conversational requests.
• Retrieves existing projects when you need to continue work.
• Lists projects and screens so you can quickly find what you already built.
• Fetches individual screens for review and iteration.
• Generates new UI screens directly from text descriptions, using Gemini-powered understanding and Stitch functions.
• Selects the right Stitch capability intelligently—the AI analyzes your request, then calls the appropriate Stitch function(s) instead of blindly hitting APIs.

Use cases

• As a SaaS operator, you want a fast way to brainstorm and iterate UI screens by messaging your design assistant in Telegram.
• As an automation engineer, you need an end-to-end n8n automation workflow that connects chat input to structured design actions in Google Stitch.
• For demos or internal prototypes, you can run design exploration via chat—creating projects, browsing screens, and generating new layouts from plain text.

Technical details

• Integrations / capabilities: Telegram for chat interaction, Google Gemini for AI interpretation, and Google Stitch (MCP) for project/screen actions.
• Workflow logic: uses nodes including if, set, code, telegram, and a sticky note for clarity.
• AI tooling: includes a Perplexity tool in the stack.
• Security / access control: Telegram execution is restricted to a specific user ID (workflow runs only when authorized).