[Suggestion] Zapier integration for audio transcription workflows (Node.js)

[deepgram-robot]

What to build

A Zapier integration (private or community app) that enables developers to trigger Deepgram transcription from Zapier workflows. Should support transcribing audio files from common triggers (file uploads, webhook events, cloud storage) and outputting structured transcript data to downstream Zaps.

Why this matters

Zapier is the largest workflow automation platform with millions of users. Non-developer teams (sales ops, customer success, content creators) need to transcribe audio files as part of automated workflows — transcribing sales call recordings, generating show notes from podcast episodes, or processing voicemail messages. A Zapier integration makes Deepgram accessible to this massive audience without requiring any code.

Suggested scope

• Language: Node.js (Zapier CLI platform)
• Deepgram APIs: STT pre-recorded (Nova-3), Audio Intelligence (summarization, topics)
• Zapier actions: Transcribe Audio File (from URL), Transcribe Audio File (from file upload)
• Zapier triggers: Transcription Complete (webhook callback)
• Output fields: Full transcript, paragraphs, summary, detected language, duration
• Complexity: Medium — requires understanding of Zapier CLI app structure

Acceptance criteria

• Runnable with minimal setup (clone, add API key, zapier test)
• README explains Zapier app structure and how to publish
• Uses current SDK version
• Supports both URL-based and file-based audio input
• Returns structured transcript data usable in downstream Zaps

---

Raised by the DX intelligence system.