Python for finance — Excel model migration, factor backtesting, close automation
The Excel workbook that crashes at month-end doesn't crash in Python. It also runs in 6 seconds instead of 6 minutes.
Every finance function eventually hits the Excel ceiling: workbook size 400MB, one wrong VLOOKUP nukes the model, the analyst who wrote it left three years ago. Python is where the escape happens. It's not about replacing Excel — pivot tables aren't going anywhere — it's about moving the load-bearing math (Monte Carlo, backtests, close automation, reconciliation) to code that can be reviewed, tested, versioned, and re-run without opening a file.
Data Science track: 104 lessons on pandas (Excel workflows → DataFrames), numpy (vectorised math), matplotlib (deck-ready charts), scipy.stats (proper distributions, not =NORM.DIST). Automation track: openpyxl for reading legacy workbooks and pandas.ExcelWriter for producing them, so nobody upstream has to change tools. Path: 90-day junior-backend if the exit is Fintech eng, or the Weekend Excel automation if the goal is 'my close takes 20 minutes not 3 days'.