Asking Questions
Chat with your documents using natural language
How to Start Chatting
- 1. Navigate to the Chat page
- 2. Start a new conversation or continue an existing one
- 3. Type your question in natural language
- 4. Hit Enter or click Send
- 5. Get instant answers with source citations!
Types of Questions
Factual Questions
Best for extracting specific information from documents
Examples:
"What is the main conclusion of this research paper?"
"What is the company's revenue for Q3 2024?"
"Who are the parties involved in this contract?"
Analytical Questions
AI synthesizes information from multiple sources
Examples:
"Compare the findings across all quarterly reports"
"What are the key differences between these two contracts?"
"Summarize the main arguments in these papers"
Clarification Questions
Get explanations in plain language
Examples:
"Explain this technical term in simple language"
"What does this clause mean in practical terms?"
"Can you elaborate on the methodology used?"
Tips for Better Answers
Be Specific
Instead of "Tell me about revenue", ask "What was the total revenue in Q3 2024?"
✓ Good
"What was the total revenue in Q3 2024?"
✗ Not Ideal
"Tell me about revenue"
Provide Context
Reference specific documents or sections when possible
✓ Good
"What does section 5 of the employment contract say about termination?"
✗ Not Ideal
"What about termination?"
Ask Follow-ups
Conversations maintain context, so you can ask follow-up questions
✓ Good
"Can you explain that in simpler terms?"
✗ Not Ideal
"Start a new conversation for every question"
Use Natural Language
Write questions as you would ask a colleague
✓ Good
"How does this compare to last year?"
✗ Not Ideal
"Compare data current year previous year"
Understanding Source Citations
Every answer includes source citations showing where the information came from:
Document Name
Which document the information came from
Page/Section
Specific location within the document
Relevance Score
How relevant this source is to your question
💡 Click on any source citation to view the original text in context
RAG vs Direct Mode
RAG Mode (Default)
Uses Retrieval-Augmented Generation to search your documents and provide answers based on your data.
Best for: Questions about your documents
Direct Mode
Directly queries the AI without searching documents. Uses the AI's general knowledge.
Best for: General questions, brainstorming