Exploratory Analysis

Use exploratory questions to understand your data, discover patterns, and identify areas for deeper investigation.

What is happening in my data?

Descriptive analysis helps you understand the basic characteristics and patterns in your dataset.

What the AI Agent Does:

• Creates bar charts, scatter plots, and distribution visualizations
• Shows data patterns and relationships
• Identifies interesting trends in your data
• Generates interactive charts you can explore

Understanding Your Dataset

• “What does this dataset contain?”
• “How many records and variables do I have?”
• “What are the data types and ranges?”
• “Are there any missing values I should know about?”

Distribution Analysis

• “What’s the distribution of [variable]?”
• “Show me the frequency of different categories”
• “How spread out are the values?”
• “Are there any unusual patterns in the data?”

What patterns and insights exist in my data?

Use exploratory questions to uncover hidden patterns and interesting relationships.

What the AI Agent Does:

• Identifies interesting patterns and trends
• Highlights outliers and unusual data points
• Creates visualizations to explore relationships
• Suggests areas for deeper investigation

Time-Based Patterns

• “How have sales changed over time?”
• “Are there seasonal patterns in the data?”
• “What trends do I see in customer behavior?”
• “When did significant changes occur?”

Segmentation Patterns

• “How do different customer segments behave?”
• “What groups exist naturally in this data?”
• “Which categories perform differently?”
• “Are there distinct clusters in the data?”

Outlier Questions

• “What are the unusual data points?”
• “Which records don’t fit the typical pattern?”
• “Are there any extreme values I should investigate?”
• “What makes certain records different?”

Anomaly Investigation

• “Why are these values so different?”
• “What happened on dates with unusual activity?”
• “Which factors contribute to outlier behavior?”
• “Should these outliers be included in analysis?”

Data Quality Questions

• “What percentage of data is missing for each variable?”
• “Are there patterns in the missing data?”
• “How consistent is the data across different sources?”
• “What data quality issues should I address?”

Validation Questions

• “Do the values make sense for each variable?”
• “Are there impossible or implausible values?”
• “How reliable is this data source?”
• “What assumptions should I verify?”

1. Begin with Overview Questions

   • “What does this dataset contain?”
   • “What are the main variables?”

2. Explore Distributions

   • “What’s the distribution of key variables?”
   • “Are there any obvious patterns?”

3. Identify Interesting Areas

   • “What looks unusual or unexpected?”
   • “Where are the potential insights?”

4. Drill Down

   • “Why does this pattern exist?”
   • “What factors contribute to this difference?”

Sequential Question Development

• Start with broad exploratory questions
• Follow up on interesting patterns you discover
• Ask more specific questions about promising areas
• Validate initial findings with additional analysis

Example Progression:

1. “What patterns exist in customer data?”
2. “Why do some customers have much higher value?”
3. “What characteristics distinguish high-value customers?”
4. “How can we identify potential high-value customers?”

• “What are the main customer behavior patterns?”
• “How do purchasing patterns vary by demographics?”
• “What seasonal trends exist in sales data?”
• “Which products are commonly bought together?”

• “How do users engage with different features?”
• “What usage patterns predict long-term retention?”
• “How does user behavior change over time?”
• “What onboarding patterns lead to success?”

• “What channels drive the most valuable traffic?”
• “How does campaign performance vary by audience?”
• “What content types generate the most engagement?”
• “How do marketing activities influence sales?”

• “What patterns exist in production efficiency?”
• “How do different processes compare in performance?”
• “What factors influence quality metrics?”
• “Where are the bottlenecks in our operations?”