Docs / Data sources / OpenAI (ChatGPT Enterprise)

OpenAI (ChatGPT Enterprise)

Business software

Built when you ask for it

Sync your ChatGPT Enterprise workspace's usage analytics, members, groups and usage limits into your warehouse. Ingest builds this connector the first time a customer asks for it, then keeps the tables current in a data warehouse you own, on the schedule you choose.

The ChatGPT Enterprise connector syncs your workspace's administration data into analytics-ready tables in your data warehouse. Pull daily ChatGPT, Codex and Work usage per member, Codex code review metrics and turn insights, the workspace's members, invitations, groups and memberships, monthly usage limits and limit requests, and service accounts from OpenAI's ChatGPT Admin API, with Ingest handling auth, pagination, and retries so the tables stay current. It uses a workspace Admin key and your Workspace ID, and reads no conversations or other content your people create. An OpenAI API Platform Admin key, which the OpenAI Admin API (API Platform) connector reads with, cannot call these endpoints.

What you would get

Each of these arrives as its own table, kept current on the schedule you choose:

  • Daily ChatGPT, Codex and Work usage
  • Codex code reviews and turn insights
  • Members, invitations and groups
  • Monthly usage limits and requests
  • Service accounts and credentials

How it gets built

  1. Choose OpenAI (ChatGPT Enterprise) when you set up a pipeline, with the warehouse it goes to and a schedule.
  2. Ingest reads OpenAI (ChatGPT Enterprise)'s documentation, and you choose the tables you want.
  3. Ingest builds the connector and tests it against OpenAI (ChatGPT Enterprise) itself.
  4. A person at Ingest reviews and publishes it, and your pipeline starts on its own.

What you will need

ChatGPT workspace Admin key (Bearer token). Only an Owner or Admin of your ChatGPT Enterprise, Edu or Healthcare workspace can create the key: in the Admin Console, open Credentials > Admin keys, select Create new admin key, and choose Read only, or Custom with read access to Users, which Ingest checks the key with, and to the data you want. Ingest also asks for your Workspace ID, the UUID on the Admin Settings page. An OpenAI API Platform Admin key cannot call these endpoints.

You enter it once, in a form in Ingest, and it is kept in a secret store set aside for your organization.

Questions

Does Ingest have a OpenAI (ChatGPT Enterprise) connector?
Not as a finished connector yet. Ingest builds it from OpenAI (ChatGPT Enterprise)'s own documentation the first time a customer asks, tests it against the real service, and lists it once it passes.
Which warehouses can OpenAI (ChatGPT Enterprise) data go to?
Amazon S3, Google Cloud Storage, Azure Blob Storage, Apache Iceberg, Ingest Managed Lakehouse, MotherDuck, Postgres, Amazon Athena, Databricks, Redshift, BigQuery, ClickHouse, MySQL / generic SQL and Snowflake, in an account you own.
Do I need to write code?
No. Ingest builds, runs and maintains the connector. Someone with access to your warehouse connects it once, following a short guide.

OpenAI (ChatGPT Enterprise) API documentation: https://learn.chatgpt.com/docs/enterprise/analytics-api

Setting up the destination is its own short guide, one per warehouse or lake: Amazon S3, Google Cloud Storage, Azure Blob Storage, Apache Iceberg, Ingest Managed Lakehouse, MotherDuck, Postgres, Amazon Athena, Databricks, Redshift, BigQuery, ClickHouse, MySQL / generic SQL, Snowflake.

Other sources Ingest connects for SaaS teams: Asana, Clockify, Google Sheets, Jira, Salesforce, Airtable.

Request OpenAI (ChatGPT Enterprise)