PlanningManager
Python OOP library that replaced a set of critical VBA macros — now maintained and extended by two other engineers.
Context
The planning department relied heavily on a suite of VBA macros embedded in Excel workbooks to process and format grid planning data. The macros were functional but impossible to test, difficult to debug, and a single point of failure — only two people understood them, and modifications required working inside Excel’s VBA editor with no version control.
Role
Solo developer. Interviewed the macro authors to document the implicit business logic, then rebuilt the functionality in Python with a clean object model.
Approach
The core design decision was to model the domain — not just replicate the macros:
- Defined a set of Python classes representing planning entities (projects, constraints, planning horizons) rather than translating the imperative macro code line by line.
- Used
openpyxlfor direct Excel file manipulation andpandasfor data transformation, preserving compatibility with existing Excel-based workflows. - Added a simple CLI interface so analysts could trigger runs from the command line without touching the code.
- Wrote inline documentation and a test suite covering the main data transformation paths.
The result was a library rather than a script — importable, testable, and extensible.
Outcome
- All VBA macros replaced without any loss of functionality.
- Two other engineers joined the codebase within the first year and have since added new features.
- Bug fix cycle reduced from “find the macro author” to a standard pull request workflow.
- The library now serves as a foundation for other automation work in the department.
What I Learned
The real work in a refactoring like this isn’t the code — it’s extracting the implicit business logic from people who’ve internalised it. The interviews were harder and more valuable than the implementation.