How to use Claude for a personal spending dashboard
Claude Cowork built Michael Saif’s monthly bank-statement spreadsheet and spending dashboard, with separate accounts and editable categories for human review.
app.therundown.ai
Suggested prompt
Suggested prompt written by Agents In Use, not the source author. Replace bracketed inputs and review tool permissions first.
Build a local Python workflow for my authorized monthly bank and card statements. Extract charges into a spreadsheet with separate account views and a combined dashboard by category and period. Keep source provenance and editable category corrections, and show parsing or reconciliation uncertainty for my review before I rely on totals.
Reach people exploring AI agents
See placements →Evidence at a glance
Sourced account- Source author
- Michael Saif
- Original account
- Read the source ↗
- Published
- Source link checked
- Reported result
- Saif reports replacing his personal QuickBooks workflow with a recurring Python, spreadsheet and dashboard process; independent accuracy and subscription cancellation proof are absent.
- Human involvement
- The owner supplies monthly source statements, can change assigned categories and reviews spending; the post does not disclose reconciliation or parsing tests.
A linked account supports attribution. It does not mean we independently reproduced its result.
Report an issue with this case
Open correction emailOpens your email app with the case link included. Nothing is sent automatically.
How the agent did it
- Saif explained the desired statement-to-spreadsheet output and Claude built the Python workflow.
- Each month he uploads bank and card statements, which the workflow parses into categorized account views.
- Saif corrects categories when needed and reviews the connected spending dashboard.
Michael Saif used Claude Cowork to build a Python workflow that replaced his personal QuickBooks setup. His recurring input is a folder of monthly bank and credit-card statements, rather than a conversational estimate of what he spent.
Saif described the desired output in a prompt. Claude built a process that parses statement charges and adds them to a spreadsheet. The spreadsheet assigns categories automatically, while leaving Saif able to change a category when it is wrong. Each bank or card account retains its own view, alongside a combined view of the household finances.
The spreadsheet connects to a dashboard for reviewing spending by category, time period and account. The account therefore describes both an implemented tool and the owner’s repeated monthly use: supplying source files, reviewing classifications and inspecting the result.
Saif says the previous personal subscription cost $38 per month. The post reports replacing that workflow, but does not publish parsing accuracy, statement formats, reconciliation tests or a cancellation receipt. Its demonstrated result is an owner-operated statement-to-dashboard process, with editable categorization as the explicit human checkpoint rather than a verified accounting or tax service.
Prompt
Suggested prompt (written by Agents In Use, not the author's)
Source: The Rundown University · Michael Saif
Frequently asked questions
Can the human correct classifications?
Yes. The author explicitly says automatically assigned categories can be changed.
Are accounts merged into one undifferentiated total?
No. Separate bank and card views coexist with a combined view.
Is the stated $38 a verified monthly saving?
It is the reported price of the replaced subscription; the post supplies no cancellation or billing receipt.
More details
About this account
Saif explained the desired statement-to-spreadsheet output and Claude built the Python workflow. Each month he uploads bank and card statements, which the workflow parses into categorized account views.
Full primary account and exact HTML publication metadata read. Monthly use and former $38 subscription price are author-reported. No original prompt text, measured accuracy or cancellation receipt was supplied. Suggested prompt written by Agents In Use; no verbatim original task prompt was supplied.
Credit: Michael Saif — The Rundown University
Source published
Added to Agents In Use
Source checked
Tools and topics
Claude Cowork · Python · Spreadsheet · Dashboard
personal statements · editable categories · account comparison
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.