Skip to Content

Not for sale yet.

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

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

Read.ai meetings, recorded in Odoo.

One endpoint at /webhook/readai/meeting_end turns Read.ai's payload into Odoo records — meeting, participants, action items, key questions, topics, chapter summaries, and a full speaker-blocked transcript.

See how it works Talk to us →

Why

Read.ai writes good summaries and great chapter breakdowns. Then it leaves them in its own portal. This module is the receiver — one webhook route, one transactional save, and every meeting lands in Odoo on the shared meeting.abstract model so the rest of your business logic can see it.

How it works

  1. Configure the Read.ai webhook
    Point your Read.ai workspace at /webhook/readai/meeting_end; the route accepts public POSTs and fires only on the meeting_end trigger.
  2. Parse the meeting envelope
    Session id, title, owner, start and end times, summary, and report URL land on a read.ai.meeting record.
  3. Fan out to child tables
    Participants, action items, key questions, and topics each get their own rows linked to the meeting.
  4. Stitch chapters to topics
    Chapter summaries resolve their topic names back to topic ids, so a chapter knows which topics belong to it.
  5. Store the transcript blocks
    Speaker and speaker-block arrays persist as JSON, then render to a coloured per-speaker transcript with partner avatars when matched.

What's in the box

Public webhook receiver

JSON-RPC route on Odoo 19, no CSRF, no auth required — gate it at your reverse proxy or behind a Read.ai shared secret you put in front.

Six linked record types

Meeting, participants, action items, key questions, topics, chapter summaries, and transcript — modelled, indexed, and exposed through the Meeting Hub portal.

Partner auto-match

Participants resolve to res.partner by email first and case-insensitive name second, so contacts pick themselves up.

Rendered HTML transcript

Speaker-coloured chat bubbles with avatar initials or partner photo, timestamps, and hover-revealed end times — readable in the portal without a Read.ai login.

Chapter-to-topic linking

Each chapter summary remembers which topics it covered, so navigation by chapter or by topic gives the same set of rows.

MCP-ready transcript shape

A normalised _get_transcript_blocks emits timestamp-second, speaker, text dicts for MCP tools and downstream LLM jobs.

Compatibility

Odoo 19.0 LGPL-3 Read.ai meeting_end webhook

Depends on meeting_hub plus Odoo base, mail, portal, calendar. Python dateutil handles the ISO datetime parsing.