AI in Finance - Ready-to-use prompts

I'm sharing a dataset exported from [ERP/system name]. Before I start working with it, help me identify:
1. Missing or blank values in key columns
2. Duplicate rows
3. Inconsistent formats (dates, currencies, text casing)
4. Any values that look like outliers or data entry errors

Here is the data: [paste data]

Give me a short summary of what you find, and for each issue suggest the fastest fix.

I have two lists that should match but don't — cost center names from our ERP and cost center names from our reporting tool. Help me:
1. Find likely matches between the two lists (fuzzy match, ignore casing/spacing)
2. Flag any names that appear in one list but not the other
3. Suggest a standardised name for each pair

List A (ERP): [paste]
List B (Reporting): [paste]

Output as a table: ERP name | Reporting name | Suggested standard name | Match confidence

I have unstructured text (email threads, meeting notes, or free-text fields) that contain financial data. Extract and structure the following fields into a table:
- Date mentioned
- Amount (with currency)
- Category or cost type
- Entity or department
- Any approval status or condition mentioned

Text: [paste]

If a field is unclear or missing, mark it as "unclear" rather than guessing.

Here is a budget vs actuals table for [month/period]:

[paste table]

For each line where variance exceeds [X]% or [€X], write a one-sentence explanation of the likely business cause. Use plain language — no jargon. If the cause is unclear, say so rather than inventing one.

Format: Line item | Budget | Actual | Variance % | Likely explanation

I want to calculate [metric, e.g. gross margin / headcount cost / revenue per customer] from first principles using business drivers.

The drivers I have available are: [list your inputs, e.g. headcount, avg salary, bonus %, employer NI rate]

Please:
1. Show the calculation logic step by step
2. Identify which drivers have the most sensitivity (if one changes 10%, what happens to the output?)
3. Write the formula in plain English, then as an Excel formula I can use

Before I send this to management, sanity check the following figures:

[paste your numbers, table, or model output]

Check for:
- Internal consistency (do subtotals add up? do ratios make sense?)
- Directional logic (if volume went up, did revenue go up too?)
- Obvious outliers or anything that would raise a question in a review meeting
- Units and currency consistency

Tell me what looks right, what looks off, and what you'd want to verify before presenting.

Write an Excel formula that does the following:

[describe in plain English what you want, e.g. "sum column C only where column A equals 'Marketing' and column B is not blank"]

Also:
- Explain what each part of the formula does in one sentence
- Warn me if there are any edge cases I should test
- Give me an alternative formula if a simpler version exists

I'm sharing a description of my current Excel model. Help me identify weaknesses and suggest improvements.

Model purpose: [e.g. monthly P&L forecast]
Structure: [e.g. inputs sheet, assumptions, outputs, charts]
Current issues I know about: [e.g. hard-coded numbers, too many circular references]

Please suggest:
1. Structural improvements (layout, separation of inputs vs calculations)
2. Formula best practices I might be missing
3. How to make it easier to hand off to someone else
4. One thing I could automate with a macro or Power Query

I do the following task manually in Excel every [week/month]:

[describe the repetitive task step by step — e.g. copy data from 3 sheets, add a timestamp column, filter rows where value > 0, paste into a summary sheet]

Suggest how I could automate this using:
1. Excel VBA (if it can stay inside Excel)
2. Python with pandas/openpyxl (if it needs more power)
3. Power Automate (if it involves file movement or email)

For each option, rate the effort to implement (low/medium/high) and give me the first 5 lines of code to get started.