Skip to main content

Explore Data using AI

Introduction​

With Explore Data mode, whether you’re exploring simple metrics or tackling complex questions, each request is broken down into logical steps to support structured analysis.

  • How do this month's numbers compare to last month? (Period Comparison)
  • What's the percent of total by category? (Percent of Total)
  • Can I track cumulative sign-ups over time? (Cumulative Totals)
  • What are our top 5 performing stores by revenue? (Top/Bottom N)

Users can see how it breaks down a question, revealing the logic behind the insights. This transparency builds trust and ensures users always understand how their data-driven decisions are formed.

holistics-ai-thinking-process

How to use​

Start with your question​

You can start with Explore data (with thinking) and can try straightforward questions:

  • Show me revenue by product category

Slice & dice with follow-up questions​

Once you have initial results, you can ask follow-up questions right in the same conversation thread:

  • Break down by country and city
  • Show me the trend of average order value over the last year
  • Compare the average order value between different categories
  • List the top 5 products by revenue
  • Compare the number of orders placed in this quarter vs the last quarter

Not sure what to ask next? Let AI guide you with helpful suggestions!

Work with AI to provide more context​

Holistics AI may occasionally miss the mark. That's expected. By asking follow-up questions or offering clarifying input, you help guide the tool to better understand your context and intent - making sure it works as a helpful sidekick instead of some too-good-to-be-true know-it-all.

Refine and reuse metrics generated by AI with GUI​

Every operation is visible and editable. If the answer isn't what you need, tweak the step in the GUI without to start over. Analysts can refine metrics and promote them for reuse in future analysis.

Supported questions​

  • Listing: "List all products launched in the last quarter"
  • Filtering: "Show transactions over $1000 from new customers"
  • Top/Bottom N: "What are our top 5 performing stores by revenue?"
  • Metric calculations: "Calculate average order value by customer segment"
  • Aggregation and breakdown: "Break down sales by category and region"
  • Percent of Total: "What percentage of total revenue comes from each product line?"
  • Period Comparison: "Compare this month's sales to the same month last year"

Coming soon​

  • Customize Chart: Make quick adjustments using natural language
  • Explain Chart: Understand trends and patterns through auto-generated insights
Let us know what you think about this document :)