Ask Your First Question

Learn how to phrase questions, understand the answer card, and interpret confidence badges.

Once you’ve connected a data source, you’re ready to ask your first question.

How to ask

Go to the Home page and type your question in the input field. Use plain, natural language — the same way you’d ask a colleague.

Good questions:

  • “What were our top 10 customers by revenue last quarter?”
  • “How many new signups did we get this week?”
  • “What’s the average deal size by region?”
  • “Show me monthly revenue for the last 12 months”

Tips for better results:

  • Be specific about time periods (“last quarter”, “this month”, “since January”)
  • Name the metric you want (“revenue”, “signups”, “bookings”)
  • Specify groupings if needed (“by region”, “by rep”, “by month”)

Understanding the answer

Every answer includes several components:

The response

A natural language summary of what the data shows, with key numbers highlighted.

The data table

A sample of the raw result rows from the query. You can see exactly what data backs the answer.

The SQL query

The generated SQL that was executed against your data source. Click to expand and inspect. Power users can edit and re-run the query directly.

Confidence badge

Every answer includes a confidence indicator:

BadgeMeaning
High (green)Reliable, well-grounded in the data
Check (amber)Looks reasonable, but review recommended
Low (red)May need refinement — check the query
Failed (red)The query couldn’t execute

Each badge includes a note explaining why that confidence level was assigned.

Metadata

  • Which data source was queried
  • Row count returned
  • Query execution time
  • Bytes processed

Follow-up questions

After getting an answer, you can refine it with follow-up questions in the same thread:

  • “Now break that down by region”
  • “Filter to just enterprise customers”
  • “Compare that to last quarter”
  • “Show me the top 5 only”

Klairr maintains the full conversation context — no need to repeat yourself.

Saving answers

Click Save as Report on any answer to preserve it. Reports can be:

  • Named and organized
  • Shared with your team
  • Scheduled for automatic refreshes
  • Exported as CSV or JSON

Feedback

Use the thumbs up/down buttons on any answer to rate it. Your feedback helps improve answer quality over time.

What’s next

Need help? Contact support · Start free