Automating Fireflies Transcripts to Outlines and Action Items: My AirOps Workflow

Описание: I use Fireflies to capture meeting transcripts and voice memos, but the data was stuck in Fireflies—not easily actionable. So I built an AirOps workflow that automatically processes those transcripts into structured outlines and action items in Google Docs.

In this video, I walk through exactly how I built this workflow and how it fits into my existing automation setup.

The problem:
Fireflies sends transcript links to Slack after meetings, but I still had to manually:

Open the Fireflies link
Copy the transcript
Paste it into ChatGPT or another tool
Format it into an outline
Identify action items
Create a doc to reference later

The solution:
An AirOps workflow that does all of this automatically.

How the workflow works:

Takes a Fireflies transcript URL as input (from Slack or anywhere)
Extracts the transcript ID using JSON formatting
Calls the Fireflies API to grab the full transcript data
Runs it through GPT-4 with a custom prompt to create:
A detailed outline with subheadings and bullet points
Action items attributed to speakers
Creates a Google Doc with all the content
Sends the doc link back to Slack for easy access

What I cover in this video:

My existing Slack + Fireflies automation
Why the transcript data wasn't useful on its own
Building the AirOps workflow step-by-step
Extracting transcript IDs via JSON formatting
Calling the Fireflies API
Creating the LLM prompt for outline generation
Easy Google Docs integration (no complicated service workers!)
Slack integration and notification setup
Creating shareable workflow links (so others can use it without logging in)
Seeing the final result in Google Docs

Use cases:
✅ Client meeting notes with action items
✅ Voice memo → blog outline conversion (my favorite!)
✅ Podcast interview summaries
✅ Any rambling transcript that needs structure

Why this workflow works for me:
I use Fireflies for both client meetings AND voice memos where I talk through blog ideas. This workflow is multi-purpose—it works for both use cases. I could refine it into separate workflows optimized for each, but the current version handles both perfectly.

Key features:

Shareable workflow links (people can run it without AirOps access)
Adaptable to other AI note-takers with API access
Simple but time-saving automation
Slack notifications keep everything organized

TIMESTAMPS:
0:00 - Intro: Augmenting existing workflows
0:17 - My Slack + Fireflies automation
1:15 - Running the AirOps workflow
1:43 - How I built the workflow (step-by-step)
2:05 - Step 1: Taking the transcript URL input
2:16 - Step 2: Extracting transcript ID with JSON
2:23 - Step 3: Calling the Fireflies API
2:58 - Step 4: LLM prompt for outline + action items
4:06 - Step 5: Google Docs integration
5:13 - Use case: Voice memos for blog outlines
6:07 - Step 6: Slack notification setup
7:03 - Seeing the final result
7:44 - Why formatting doesn't matter to me

Have you automated your transcripts? What's your workflow? Let me know in the comments!