Ingestion Methods

Ingestion methods are the different ways you can add documents to a collection in Enterprise h2oGPTe. Whether your content lives on your computer, on the web, in cloud storage, or in another collection, there's a method to bring it in.

This section explains each option and helps you pick the right one for your source.

Every ingestion method adds documents to the same collection and applies that collection's parsing and PII rules. They differ mainly in where the content comes from and how you authenticate to reach it: local uploads need no setup, while remote and enterprise connectors require credentials. Connectors such as S3, Azure Blob Storage, Google Cloud Storage, SharePoint Online, SharePoint On-Premise, and Confluence also support scheduled auto-sync.

Choosing a method

Use this matrix to quickly match your source to the right ingestion method.

Method	Best for	Source type	Additional resource
Upload Documents	Quickly adding a few local files by drag-and-drop or selection	Local files	Upload Documents
Import from File System	Bulk imports from directories using glob patterns	Local or network directories	Import from File System
Import from URL	Web pages, articles, and crawled documentation sites	Web URLs	Import from URL
Upload Plain Text	Notes, transcripts, and code snippets entered or pasted directly	Pasted text	Upload Plain Text
Select a Document	Reusing a document from another collection, re-parsed with current rules	Existing document	Select a Document
Select a Collection	Merging or copying every document from another collection	Existing collection	Select a Collection
Import from S3	Documents stored in Amazon S3 buckets	Amazon S3	Import from S3
Import from Azure Blob Storage	Documents stored in Azure Blob Storage containers	Azure Blob Storage	Import from Azure Blob Storage
Import from Google Cloud Storage	Documents stored in Google Cloud Storage buckets	Google Cloud Storage	Import from Google Cloud Storage
Import from SharePoint Online	SharePoint Online sites and document libraries	Microsoft 365	Import from SharePoint Online
Import from SharePoint On-Premise	On-premises SharePoint installations	On-premises SharePoint	Import from SharePoint On-Premise
Import from Confluence	Confluence Cloud pages, spaces, and attachments	Atlassian Confluence	Import from Confluence

In this section

upload_file

Upload Documents

Upload files directly from your computer with drag-and-drop or file selection. The most straightforward way to add a few local documents.

folder_open

Import from File System

Import documents from local or network directories using glob patterns. Ideal for bulk imports from organized folder structures.

link

Import from URL

Import content from web URLs with configurable crawling. Crawl websites and extract articles, documentation, and other online content.

text_snippet

Upload Plain Text

Create documents by entering or pasting text directly. Perfect for notes, transcripts, code snippets, and content that isn't a file.

file_copy

Select a Document

Reuse a document from another collection. The original file is re-parsed with the current collection's parsing and PII rules.

library_books

Select a Collection

Import every document from another collection into your current one. Useful for merging, copying, or templating collections.

cloud

Import from S3

Import documents from Amazon S3 buckets with configurable credentials and paths. Supports IAM roles and auto-sync.

cloud_queue

Import from Azure Blob Storage

Import documents from Azure Blob Storage containers using account keys, SAS tokens, connection strings, or Managed Identity, with auto-sync.

cloud_done

Import from Google Cloud Storage

Import documents from Google Cloud Storage buckets using service accounts, Workload Identity, or ADC, with auto-sync support.

groups

Import from SharePoint Online

Import from SharePoint Online sites and document libraries through the Microsoft Graph API with modern authentication and auto-sync.

dns

Import from SharePoint On-Premise

Import from on-premises SharePoint installations using legacy authentication methods such as Windows and claims-based auth.

article

Import from Confluence

Import pages, spaces, sub-pages, and embedded images from Atlassian Confluence Cloud using OAuth or API token authentication.

Next steps

• New to collections? Start with Create a collection, then return here to add documents.
• For shared processing settings that apply across methods, see Add documents to a collection.
• To keep cloud and enterprise sources up to date automatically, see Auto-Sync Connectors.

---

Feedback

• Submit and view feedback for this page
• Send feedback about Enterprise h2oGPTe to cloud-feedback@h2o.ai