Skip to Content

Not for sale yet.

Drop your email and we'll let you know the moment it goes live.

Fireflies.ai Integration · Harrison Consulting
ODOO MODULE · MEETING INTELLIGENCE

Fireflies transcripts, landed in Odoo.

A signed webhook catches each finished transcript, GraphQL fetches the rest, and a nightly cron backfills anything the webhook missed. Records live on the shared meeting.abstract model.

See how it works Talk to us →

Why

Fireflies is fast at recording and slow at being part of the rest of your business. Action items sit in a portal nobody opens. This module pulls every transcript into Odoo the moment Fireflies finishes processing, then keeps the rest in sync on a daily schedule so missed webhooks don't become missing meetings.

How it works

  1. Webhook on transcription complete
    Point Fireflies at /webhook/fireflies/transcription_complete; the controller verifies the x-hub-signature HMAC-SHA256 against your configured secret.
  2. GraphQL fetch of the transcript
    A single transcript(id) query pulls title, dates, summary, sentences, speakers, attendees, soundbites, and audio/video URLs.
  3. Records persisted across nine tables
    Meeting, participants, speakers, sentences, topics, action items, sentence-level actions and questions, and tags — all linked, all indexed.
  4. Daily backfill cron
    Runs at 02:00 to catch meetings the webhook missed and refresh any that arrived without a summary yet.
  5. Surfaces in Meeting Hub portal
    Records appear at /my/meetings next to Read.ai and Voice Bot sessions, filtered by portal access rules.

What's in the box

Signed webhook endpoint

HMAC-SHA256 verification against a system parameter secret; duplicate meetings are short-circuited before any work happens.

GraphQL sync service

One service handles webhook fetches, manual refreshes, batch backfill, and a Fireflies uploadAudio mutation for sending audio in.

Speaker analytics

Per-speaker talk time, words-per-minute, longest monologue, and sentiment counts computed from the sentence stream.

Sentiment scoring

Positive, negative, and neutral sentence counts roll up to a polarity index on both the meeting and each speaker.

Action item extraction

Summary-level action items plus sentence-level action items and questions, each pinned to the speaker and timestamp that produced them.

Add-to-live wizard

A backend wizard sends the Fireflies bot into a live meeting URL with title, password, duration, and language picked at submission time.

Compatibility

Odoo 19.0 LGPL-3 Fireflies GraphQL API

Depends on meeting_hub, queue_job, queue_preflight, plus Odoo calendar, portal, mail. Python requests and python-dateutil are required.