Automated Meeting Transcription with AssemblyAI

Automatically transcribes meeting audio or video with AssemblyAI, tracking status and saving transcript metadata for downstream use.

Created by Chris Moen • Version 1 • 9 steps

Streamline Your Meeting Documentation

This workflow automates the entire lifecycle of meeting transcription, from initial file upload to long-term storage. By leveraging the AssemblyAI API, the flow monitors processing status, extracts speech-to-text data, and organizes metadata without manual intervention. It eliminates the need to manually check for finished jobs, ensuring your transcripts are ready as soon as the processing is complete.

Robust Data Management and Indexing

Designed for efficiency, this template uses a multi-step logic to manage transcription records. It automatically builds daily indexes and stores transcript records for 30 days using versatile Key-Value (KV) storage. The flow efficiently handles:

  • Automated Job Submission: Converts audio or video files into transcription requests.
  • Status Tracking: Continually polls the AssemblyAI server to confirm completion.
  • Metadata Organization: Structures and saves transcript data for easy retrieval by downstream applications.

Key Benefits of This Automation

  • Scalable Speech-to-Text: Handle high volumes of meeting recordings with consistent accuracy using advanced AI.
  • Seamless Integrations: Connects AssemblyAI via HTTP requests with internal function nodes for custom data shaping.
  • Searchable Archives: Automated indexing ensures you can find specific meeting records quickly.
  • Developer Friendly: Easy to customize the retention period or add steps like sending summaries to Slack or email.

Steps

  1. Check AssemblyAI transcription status (http)
  2. Load saved transcription metadata (kv)
  3. Build completed transcript record (function)
  4. Save transcription status for 30 days (kv)
  5. Save completed transcript record (kv)
  6. Add completed transcript to daily index (kv)
  7. Build transcription jobs (function)
  8. Start AssemblyAI transcription (http)
  9. Save AssemblyAI transcription metadata (kv)

FAQ

What does the Automated Meeting Transcription flow do?

This workflow automates the process of converting meeting audio or video into text using AssemblyAI. It handles the entire lifecycle of a transcription job, from submission and status tracking to saving the final metadata for organizational use.

Which integrations are required for this transcription workflow?

The flow utilizes a combination of AssemblyAI for speech-to-text processing and a Key-Value (KV) store to manage metadata and daily indices. It also uses HTTP steps to communicate with the AssemblyAI API and internal functions to structure the transcript data.

How does the flow handle transcription status and storage?

The workflow uses an automated status check to monitor the progress of AssemblyAI jobs. Once a recording is processed, it builds a completed transcript record, saves it to storage for 30 days, and automatically updates a daily index for easy retrieval.

Can I customize where the final meeting transcripts are saved?

Yes, because the flow is built using modular HTTP and function steps, you can easily extend it to send the finished transcripts to other downstream apps like Slack, Notion, or a CRM once the transcription status is marked as complete.