Practice Prompts & Clean Data

Download realistic mock datasets and complete these exercises to build your portfolio. All data is entirely fictional and safe to use publicly.

Can't Scrub Your Real Data?

If you don't have time to scrub sensitive information from your actual work files—or you simply can't share them due to confidentiality agreements—these practice prompts provide pre-generated clean datasets you can use instead. Download the CSV files, complete the exercise in Excel/Python/SQL, and upload your solution to your portfolio.

The Broken Bank Rec
AccountingIntermediate

The Scenario

A high-volume e-commerce client has a mismatch between their Shopify payout report and their Chase bank statement.

The Data

Two CSVs: One with 500 bank transactions, one with 510 internal ledger entries (including 10 duplicates/missing entries).

The Goal

Upload a reconciliation template that identifies the specific 'reconciling items' and provides the adjusting journal entry.

Fixed Asset Roll-Forward
Core AccountingIntermediate

The Scenario

A manufacturing firm has acquired 15 new pieces of equipment mid-year with different depreciation lives.

The Data

A 'Beginning Balance' list and an 'Additions/Disposals' log with dates, costs, and asset classes.

The Goal

Create a dynamic Roll-Forward schedule in Excel that calculates YTD Depreciation and Ending Net Book Value.

Download Clean Data

SaaS Rule of 40 Dashboard
FP&AIntermediate

The Scenario

A Series B Fintech startup needs to see if they are 'healthy' according to the Rule of 40.

The Data

24 months of P&L data (Revenue, OPEX, Net Income).

The Goal

A visualization (Excel or BI) showing Revenue Growth % + Profit Margin % over time, highlighting where they hit or miss the '40' mark.

Download Clean Data

Python: Automated Intercompany Elimination
Fintech/OpsAdvanced

The Scenario

A global company has 5 subsidiaries charging each other 'Management Fees' that must be eliminated at the TopCo level.

The Data

A combined Trial Balance with 5,000 rows, including 'Due To/From' accounts with differing entity codes.

The Goal

A .py script that identifies matching intercompany pairs and generates a 'Clean' Consolidated TB.

Download Clean Data

Loan Book Stress Test
Fintech/LendingAdvanced

The Scenario

A 'Buy Now, Pay Later' (BNPL) firm wants to see the impact of a 2% rise in default rates.

The Data

A portfolio of 1,000 active loans with balances, interest rates, and current 'Days Past Due' (DPD).

The Goal

A sensitivity analysis showing the impact on Net Interest Margin (NIM) and Cash Flow.

Download Clean Data

The Messy Vendor Master Cleanup
Audit/OpsIntermediate

The Scenario

An AP department has 'Apple Inc,' 'Apple,' and 'Apple, Inc.' in their system, causing double payments.

The Data

A vendor list of 2,000 names with addresses and tax IDs.

The Goal

Use Fuzzy Lookup (Excel) or a Python script to deduplicate the list and identify potential overpayments.

Download Clean Data

Three-Statement Model Build
Investment Banking/FP&AAdvanced

The Scenario

Project the next 3 years for a subscription-based business.

The Data

Last year's Balance Sheet and a list of assumptions (Churn rate, CAC, Payroll growth).

The Goal

An integrated 3-statement model where the Balance Sheet balances automatically based on the projections.

Download Clean Data

SQL: Revenue Leakage Detection
Fintech/DataIntermediate

The Scenario

Some users are getting 'Premium' features without a 'Paid' subscription status in the billing table.

The Data

Two tables: User_Activity (Feature logs) and Subscriptions (Payment status).

The Goal

A SQL query that joins these tables to find 'Feature Use' where Payment_Status != 'Active'.

Unit Economics (CAC/LTV) Analysis
Fintech/Marketing FinanceIntermediate

The Scenario

Management wants to know if their Instagram ad spend is actually profitable.

The Data

Ad Spend by month, New Customers acquired, and Average Monthly Revenue per user.

The Goal

A cohort analysis showing the 'Months to Payback' for each monthly marketing spend.

Download Clean Data

Crypto-to-Fiat Tax Reconciliation
Fintech/TaxAdvanced

The Scenario

A user has 100 trades between BTC, ETH, and USD.

The Data

A raw JSON or CSV export from an exchange API with timestamps and 'Gas Fees.'

The Goal

A schedule calculating the Realized Gain/Loss using the FIFO (First-In-First-Out) method.

Download Clean Data

Ready to Upload Your Work?

Once you've completed an exercise, head to your Portfolio page to upload your solution.