Member-only story
Coding An End-To-End Credit Card Spend Pipeline In 1 Weekend
Visualizing 48 Months Of Credit Card Spending With PyPDF, SQL & Looker (Part II).
Since the last installment of “Zach does way too much work to fix a vendor’s minor shortcoming” focused heavily on rationale, problem context and the act of processing PDFs, I want to dedicate this piece to taking all the pieces of pt I to create an actionable dashboard.
In popular terms, this edition will “make it [execution] make sense.”
Fun fact: The end product of this dashboard is what drove my wife and I to have a retrospective discussion about spending in 2024, something we’ve never done before because we weren’t equipped with the tools to examine our credit cards as a whole entity.
But before combining everything, it’s necessary to take all the raw totals and the few instances where I’m using a SUM() of transaction data and build the sources for the dash.
Aggregating Credit Card Data Into 1 View
Since this was really a “quick and dirty” exercise over a holiday break, I didn’t think too deeply about design for source data. I used four separate tables for each card:
- Chase
- American Express