---
title: "Data Administration"
description: "Administer data sources, Outposts, and data access"
canonical_url: "https://meingpt.com/en/docs/admin-guide/data-administration"
language: en
---

# Data Administration

This area controls data onboarding for assistants and RAG-based usage.

## Core functions

- create and manage data sources
- monitor data pool status, sync, and re-auth
- manage team access to data sources
- manage Outposts (feature-dependent)

## Deletion concept, retention, and data subject rights

- **Individual chats**: users can delete their own chats at any time through the interface; complete deletion including backups typically finishes within 7 days. Details: [Data flow in meinGPT chats](/en/docs/privacy-security/faq-chats-datenfluss).
- **Automatic, time-based deletion**: there is currently no workspace-wide setting that automatically deletes chats after a configurable period.
- **Per-user retention period**: there is currently no setting that lets an individual user configure their own retention period.
- **Bulk deletion**: there is currently no function to delete multiple chats at once - chats are deleted one at a time.
- **When someone leaves the company**: neither manually removing a member ([Member Management](/en/docs/admin-guide/member-management)) nor SCIM-based deactivation ([Entra ID SCIM provisioning](/en/docs/admin-guide/entra-scim-provisioning)) deletes the account or its chats - both deactivate the membership reversibly, and the data is retained.

## Admin practice

- define access and teams first, then connect sources
- monitor source sync and health regularly

## Technical Deep Dive

- [Data Pools (RAG)](/en/docs/integrations/data-pools-rag)
- [Data Sources](/en/docs/integrations/vault/sources)
- [Outpost (on Premise)](/en/docs/integrations/vault)
- [Custom Data Preparation Pipelines](/en/docs/integrations/custom-data-preparation-pipelines)