Getting Started
AnswerVault makes it easy to chat with your organization's documents. AI is built in, so there's nothing extra to configure. Follow these two steps to get started.
Step 1: Connect your documents
The first step is to connect AnswerVault to your document repositories. We support multiple platforms so you can choose where your documents live.
Supported platforms
- SharePoint Online - Connect to your Microsoft SharePoint document libraries
- Google Drive - Index documents stored in Google Drive
- Confluence - Connect to Atlassian Confluence for documentation systems
How to connect
- Log in to your AnswerVault account
- Navigate to Connections
- Click Connect
- Select your platform (SharePoint, Google Drive, or Confluence)
- Authorize AnswerVault to access your documents using OAuth
- Select which folders or spaces you want to index
- Click Start Indexing
β οΈ Important: The connection uses your existing access, so any files you can see will be available to select for syncing. Please be careful not to include confidential content you donβt intend to share.
π‘ Tip: You can connect multiple data sources from different platforms. AnswerVault will index all of them and create a unified knowledge base.
Once you've connected your data sources, AnswerVault will automatically index your documents and build a knowledge graph. This usually takes a few minutes depending on the number and size of your documents.
What gets indexed?
AnswerVault indexes:
- Document content (PDFs, Word documents, Confluence pages, etc.)
- File metadata (title, author, last modified date)
- Document structure and relationships
- Comments and discussions (where available)
Step 2: Start chatting
You're ready! Now you can start asking questions about your documents. AnswerVault supports multiple interfaces so you can use it wherever you work.
Available interfaces
- Web Interface - Available now at
app.answervault.ai
- Microsoft Teams - Available on Pro tier or higher
- Slack - Available on Pro tier or higher
- CLI - Available for developers and automation
- REST API - Available for custom integrations
Web Interface
The easiest way to get started is with the web interface:
- Log in to AnswerVault at
app.answervault.ai
- You'll see the chat interface with a message box at the bottom
- Type your question in natural language
- AnswerVault will search your documents and generate an answer
- Each response includes citations showing which documents were used
Example questions
Try asking questions like:
- "What's our data retention policy for client records?"
- "What are the access control requirements for the secure zone?"
- "Summarise the findings from the last internal audit"
- "What's the escalation process for a data breach?"
- "Who's responsible for GDPR compliance in the EU subsidiary?"
CLI Usage
If you prefer the command line, download and install the AnswerVault CLI:
curl -fsSL https://answervault.ai/cli/0.0.5/ka-darwin-arm64 -o ka
chmod +x ka
sudo mv ka /usr/local/bin/
See the CLI page for macOS Intel and Linux downloads.
Then authenticate and start asking questions:
ka auth login
ka chat "What's in the company handbook?"
Teams Integration
To add AnswerVault to Microsoft Teams:
- Go to Settings β Integrations β Teams
- Click Add to Teams
- Approve the app installation in Teams
- In any Teams channel, mention @AnswerVault with your question
Slack Integration
To add AnswerVault to Slack:
- Go to Settings β Integrations β Slack
- Click Install to Slack
- Authorize the app in Slack
- In any Slack channel, mention
@AnswerVault followed by your question
β¨ Pro tip: For best results, ask specific questions rather than broad ones. "What's our data retention policy?" works better than "Tell me about policies".
Next steps
- Configure access controls - Set up user permissions so people only see documents they should have access to
- Train your team - Share best practices for asking questions effectively
- Monitor usage - Check the analytics dashboard to see how AnswerVault is being used
- Optimize your knowledge base - Review which documents are most valuable and ensure you're indexing the right content