What to build
A working demo showing how to set up and use the Deepgram MCP server (deepgram-mcp) with AI coding assistants like Claude Desktop, Cursor, and Windsurf. The demo should include a sample project where the MCP server is configured to transcribe audio files, analyze meeting recordings, and extract insights — all from within an AI coding workflow. Include example prompts and a pre-configured MCP settings file.
Why this matters
AI coding assistants are becoming the primary development interface for many developers. MCP (Model Context Protocol) servers allow these assistants to use external tools. Deepgram's MCP server is published on PyPI and has a superior proxy architecture (tools are fetched at runtime, so new capabilities appear without client upgrades), but developers don't know it exists or how to use it. This example bridges the gap between "Deepgram has an MCP server" and "here's how I use it in my daily coding workflow" — transcribing recorded meetings for context, analyzing audio feedback, or extracting action items from voice memos while coding.
Suggested scope
- Language: Python (MCP server) + configuration files
- Platforms: Claude Desktop, Cursor, Windsurf (MCP settings examples for each)
- Deepgram APIs used: Pre-recorded STT, Audio Intelligence (summarization, topics)
- Complexity: Low — primarily configuration and usage documentation
- Include sample
claude_desktop_config.json / .cursor/mcp.json with Deepgram MCP settings
- Include 3-5 example prompts showing practical use cases
- Include a sample audio file for testing
- Show both
pip install deepgram-mcp and dg mcp CLI approaches
Acceptance criteria
Raised by the DX intelligence system.
What to build
A working demo showing how to set up and use the Deepgram MCP server (
deepgram-mcp) with AI coding assistants like Claude Desktop, Cursor, and Windsurf. The demo should include a sample project where the MCP server is configured to transcribe audio files, analyze meeting recordings, and extract insights — all from within an AI coding workflow. Include example prompts and a pre-configured MCP settings file.Why this matters
AI coding assistants are becoming the primary development interface for many developers. MCP (Model Context Protocol) servers allow these assistants to use external tools. Deepgram's MCP server is published on PyPI and has a superior proxy architecture (tools are fetched at runtime, so new capabilities appear without client upgrades), but developers don't know it exists or how to use it. This example bridges the gap between "Deepgram has an MCP server" and "here's how I use it in my daily coding workflow" — transcribing recorded meetings for context, analyzing audio feedback, or extracting action items from voice memos while coding.
Suggested scope
claude_desktop_config.json/.cursor/mcp.jsonwith Deepgram MCP settingspip install deepgram-mcpanddg mcpCLI approachesAcceptance criteria
Raised by the DX intelligence system.