Turn Japanese meeting notes into verified action items with OpenAI and data tables

Quick overview

This workflow collects Japanese meeting notes via an n8n form, uses OpenAI to extract and then verify action items against the original notes and a team list, saves the verified tasks to n8n Data Tables, and shows a results page with tasks, decisions, follow-ups, and removals.

How it works

1. Receives meeting name, meeting date, and Japanese meeting notes from an n8n Form Trigger submission.
2. Loads the team member list (names, aliases, emails) from the n8n team_members data table and compiles it for later matching.
3. Uses OpenAI to extract draft action items (task, owner, due date, evidence quote) and a list of meeting decisions from the notes.
4. Uses OpenAI again to verify the draft items against the original notes and team list, removing unsupported or duplicate items, normalizing owners, and producing follow-up questions and removal reasons.
5. Matches verified owners to the team list to fill in owner emails, sets defaults for missing owners/due dates, and inserts the resulting rows into the n8n meeting_tasks data table when tasks exist.
6. Displays a completion page showing the saved tasks plus the extracted decisions, follow-ups, and items removed during verification.

Setup

1. Add an OpenAI credential and select a chat model in the OpenAI Chat Model node.
2. Create two n8n Data Tables named team_members (name, aliases, email) and meeting_tasks (meeting, meeting_date, task, owner, owner_email, due, evidence, status).
3. Populate team_members with your team (optionally run the sample-team branch once, then edit/remove sample entries to avoid duplicates).
4. Open the form produced by the Form Trigger and share its URL with users who will paste meeting notes.