AI Prompts
AI data analysis prompts.
Ready-to-use prompts for exploring, cleaning, querying and reporting on data — tap to copy, then fill in the blanks.
Tap Copy on any prompt, paste it into your AI assistant, and replace the [bracketed] parts with your own details.
Explore & summarize
First look at a dataset
Act as a data analyst. Here's a sample of my dataset in CSV form: [paste rows with headers]. Summarize what it contains, list the columns and their likely types, and give me three first-glance observations worth investigating.
Find the story
Based on this data, identify the 3-5 most interesting trends, patterns or outliers. For each, explain what it suggests and one follow-up question I should ask: [paste the data].
Descriptive stats
Calculate descriptive statistics (mean, median, min, max, standard deviation) for each numeric column below, and flag any column with unusually high spread or obvious outliers: [paste the data].
Clean & prepare
Data-quality check
Perform a data-quality check on this table. Tell me the type of each column, where there are missing or null values, and any likely duplicates or inconsistencies: [paste the data].
Cleaning script
Write a pandas script to clean this dataset: fix inconsistent headers, handle missing values sensibly, remove duplicates, and standardize date formats. Explain each step briefly. Columns: [describe your columns].
Reshape the data
I need to turn this [wide/long] data into [long/wide] format for analysis, keyed by [column]. Give me the pandas code and show what the result should look like: [paste a sample].
Query & code
Write a SQL query
Write a SQL query for this question: [your question, e.g. the top 5 customers by total spend last year]. Tables and key columns: [list them]. Explain the joins and grouping in one line.
Explain a query
Explain what this SQL does in plain English, step by step, and point out anything that looks inefficient or risky: [paste the SQL].
Build a formula
Give me an Excel / Google Sheets formula to [do X, e.g. count rows where region is West and sales exceed 1000] for data in [range], and explain how it works.
Explain & report
Executive summary
Turn these findings into a crisp executive summary for non-technical leaders: 3-4 bullet takeaways, no jargon, each tied to a decision or action. Findings: [paste your analysis].
Suggest a chart
Recommend the best chart type to communicate [what you want to show] from this data, explain why, and suggest what goes on each axis: [describe or paste the data].
Peer-review my analysis
Act as a skeptical reviewer. Here's my summary of findings: [paste it]. What am I likely missing (confounders, selection bias, seasonality, sample-size issues)? List each concern and how to check it.
From raw rows to real answers
AI is genuinely useful for the tedious middle of analysis — profiling a dataset, writing cleaning and query code, and turning findings into plain-English summaries. Paste a small, representative sample with headers rather than a huge file, and never paste confidential data. For spreadsheet-specific help see Excel prompts, and for writing the surrounding code, coding prompts.
FAQ
Is it safe to paste my data into AI?
Don't paste confidential, personal or proprietary data. Use a small anonymized sample or a made-up example with the same structure, then apply the code or approach to your real data yourself.
Can AI analyze a whole spreadsheet?
It works best on a representative sample or a summary rather than a huge file, because of context limits. For heavy lifting, ask it to write the pandas or SQL that you then run on the full dataset.