Transaction Categorization Assistant

Marcus, the Digital Banking Product Manager at Monzo, has a problem that's keeping him up at night. Their customers are constantly calling support asking, "What was that $47.23 charge from 'SQ *COFFEE CORNER NYC' anyway?" or "Why do I have so many 'Miscellaneous' transactions in my spending report?"

The raw transaction data from payment processors looks like hieroglyphics: "AMZN MKTP US*TO4A23EF5", "SQ *DOWNTOWN YOGA", "PAYPAL *SPOTIFY". Customers see these cryptic descriptions in their banking app and get frustrated because they can't track their spending patterns or understand where their money actually goes.

Marcus knows that if they could automatically transform these mysterious merchant codes into friendly categories like "Shopping", "Food & Dining", or "Entertainment", customers would love their banking app and call support 60% less often. The technology exists - they just need someone to build it.

Project Details

Build a service that takes raw transaction descriptions and uses a simple classification model to assign user-friendly categories


Difficulty

intermediate

Example Business

Monzo (financial/banking)

Categories

Ai Automation

Tags

transactions
classification
machine-learning
api

© 2025 Sliced Logic. All rights reserved.

Transaction Categorization Assistant