Analyze Data with AI: From Raw Data to Insights

I have a dataset I need to analyze. Let me describe it before we begin.

**Dataset name:** {{dataset_name}}
**Source:** {{where_the_data_came_from}}
**Time period:** {{date_range}}
**Number of rows:** {{approximate_row_count}}

**Columns:**
{{list_each_column_with_type_and_description}}

**Context:** {{what_this_data_is_used_for_and_why_you_are_analyzing_it}}

**Sample data (first 5 rows):**
{{paste_sample_rows}}

Based on this description, please:
1. Confirm your understanding of the dataset
2. Identify any potential data quality issues you notice in the sample
3. Suggest 5 analytical questions that would be most valuable to explore given the business context
4. Recommend what summary statistics would be useful to calculate first

Using the dataset I described, perform a comparative analysis:

**Compare:** {{metric_to_compare}} (e.g., revenue, conversion rate, response time)
**Across:** {{dimension}} (e.g., time periods, regions, product categories, customer segments)
**Focus on:** {{what_you_care_about}} (e.g., growth trends, outliers, biggest differences)

Please provide:
1. A summary table showing the comparison
2. The key finding in one sentence
3. Possible explanations for the patterns you see
4. What additional data would help confirm or disprove these explanations

Use specific numbers from the data — do not make vague statements like "there was some growth." Quantify everything.

Look at the dataset I described and identify anomalies — data points that deviate significantly from the expected pattern.

For each anomaly, provide:
1. **What:** The specific data point or pattern that is unusual
2. **How unusual:** Quantify the deviation (e.g., "3x the average", "2 standard deviations above the mean")
3. **When:** The time period or condition when the anomaly occurred
4. **Possible causes:** 2-3 hypotheses for why this anomaly exists
5. **Action:** What should be investigated or done about it

Sort anomalies by business impact (highest impact first). Ignore statistical noise — only flag anomalies that could affect business decisions.

Based on everything we have analyzed so far, synthesize the findings into a structured insights report.

**Format:**

## Top 3 Insights (ranked by business impact)
For each insight:
- **Finding:** One clear sentence stating the insight
- **Evidence:** The specific data points that support this
- **Impact:** What this means for the business in concrete terms (revenue, costs, customer satisfaction)
- **Recommended action:** One specific thing to do next
- **Confidence level:** High / Medium / Low — and what would increase your confidence

## Surprising Findings
Anything that contradicts common assumptions or expectations

## Data Gaps
What additional data would make this analysis more complete

## Executive Summary
A 3-sentence summary suitable for sharing with leadership

Be direct and specific. Replace qualitative language ("significant growth") with quantitative language ("23% increase over 6 months").

Based on our analysis, recommend the best visualizations to communicate the key findings.

For each visualization, provide:
1. **Chart type** (bar, line, scatter, heatmap, etc.) and why this type works best for this data
2. **X-axis:** What variable goes on the horizontal axis
3. **Y-axis:** What variable goes on the vertical axis
4. **Color/grouping:** How to segment the data visually
5. **Title:** A clear title that states the insight (not just "Revenue by Month" but "Revenue Grew 23% After Campaign Launch")
6. **Key annotation:** One callout or annotation to add that highlights the most important finding

Recommend 3-4 visualizations total:
- One overview chart that shows the big picture
- One detail chart that supports the most important insight
- One comparison chart that shows before/after or segment differences
- One optional chart for a surprising or secondary finding

If I am using {{tool}} (e.g., Excel, Google Sheets, Python matplotlib), provide the specific steps or code to create each chart.