This video will walk you through TeamAI's powerful data storage features that allow you to chat directly with your documents, whether for quick, one-time analysis or as part of organized collections for long-term use.
You'll learn how to upload individual documents for immediate use, create structured data collections for ongoing projects, and even integrate website content that updates automatically. By the end of this tutorial, you'll be able to efficiently manage your document library and extract valuable insights using AI-powered document chat.
Materials and Requirements
A TeamAI account with workspace access
Documents you want to analyze (PDFs, text files)
Website URLs (if you plan to import website content)
Administrative privileges (for creating collections, optional)
Step-by-Step Instructions
Chatting with Documents Already in Datastores
If your team has already set up Datastores with documents, accessing them is straightforward:
Navigate to your chat window
Select "Internal" chat mode instead of web or research modes
The data controls panel will appear on the right sidebar
Choose a specific collection if needed, or leave unselected to search all documents
Type your query, and TeamAI will search your document library for relevant information
Pro Tip: For large document libraries with thousands of files, narrow your search by selecting specific collections related to your query.
The system will perform an internal knowledge base search and return results directly from your uploaded documents, complete with source citations.
Uploading Documents for Single-Use Chat
When you need to quickly analyze documents without storing them permanently:
In the main chat window, click the plus (+) sign
Select and upload your PDF or document files
Ask questions about the document content
TeamAI will process and analyze the files for you
This method is perfect for one-time document analysis when you don't need to store files for future use or team sharing.
Creating Data Collections and Datastores
For long-term document management that can be accessed across projects and team members:
Setting Up a Collection (Top-Level Organization)
Click on "Data Storage" in the left sidebar
Click "Create Collection"
Name your collection (consider organizing by project, team, or client)
Add a descriptive note to help others understand its purpose
Set visibility (workspace-specific or organization-wide for enterprise plans)
Click "Create"
Important Consideration: If you have many team members, use clear naming conventions and descriptions so others can easily find the right collections.
After creating your collection, you'll need to create Datastores within it:
Creating a Datastore
In your new collection, click "Create Datastore"
Name your Datastore (e.g., "AI PDFs," "Project Documentation," etc.)
Click "Create"
Datastores function as sub-buckets within your main collection, allowing for better organization of related documents.
Uploading Documents to Datastores
Once your Datastore is created:
Click the "Upload" button within your Datastore
Select multiple files to upload at once
Name the document group
Choose whether to keep files grouped or split them into separate Datastores
Click "Create Document"
The system will process your uploads, which may take a moment depending on file size and quantity.
Using Stored Documents in Chat
To access your newly uploaded documents:
Go to your chat interface
Select "Internal" chat mode
On the right sidebar, choose your collection and Datastore
Ask your question
TeamAI will search your selected documents and provide relevant information with citations
This method saves time for documents you reference frequently or need to share with teammates.
Adding Website Content to Datastores
TeamAI can also import and index website content:
Adding Web Pages or Entire Websites
In your Datastore, click "Create Document"
Choose either "Web Pages" to specify individual URLs or "Website" to crawl an entire domain
For websites, enter the domain name and click "Create"
The system will crawl the website and organize content by section (products, services, articles, etc.)
Website content will automatically sync once a month
This feature is particularly useful for technical documentation, help centers, or any online resources you reference regularly.
Connecting Datastores to Agents
For advanced use cases, you can connect your Datastores to AI agents:
When creating an agent, reach the "Connect Data Hub" step
Select the relevant data collections or stores
Your agent will now have access to this information when responding to queries
This allows you to create specialized assistants with knowledge of your specific documents, perfect for customer service, technical support, or sales teams.
Tips and Warnings
Tips:
Organize collections logically by project, client, or purpose to make navigation easier
Use descriptive names and add detailed descriptions for your collections and Datastores
For large document libraries, use specific collection selections to narrow searches
Set up website syncing for documentation you reference frequently that may update over time
Warnings:
Uploading the same documents repeatedly for single chats is inefficient; use Datastores for documents you'll reference multiple times
Without proper organization, large document libraries can become difficult to navigate
Consider access permissions carefully when setting up collections shared across workspaces
Conclusion
TeamAI's Datastores and collections provide a powerful way to organize, access, and interact with your documents. By following this guide, you can efficiently manage your document library, from quick one-time document analysis to sophisticated long-term document management across teams and projects.
The ability to chat with your documents, whether they're PDFs, text files, or website content, transforms information retrieval and makes your team more efficient.
FAQs
Q: How often does website content update? A: Website content syncs automatically once a month to keep your information current.
Q: Can I chat with specific sections of my document library? A: Yes, by selecting specific collections or Datastores from the sidebar, you can narrow your search to relevant document sets.
Q: What file types are supported? A: TeamAI supports PDFs, plain text documents, and can import website content directly.