Software Contractor (Consultant) on a roughly August 2024 to April 2025 contract, about 20 hours a week, part-time alongside my other positions.
The work
Reviewed the MySQL database structure and documented the schema across its tables and relationships
Identified unused and outdated keys and table fields, stating per item whether it could be removed or whether a report or job still touched it
Produced a database diagram with DBeaver showing the tables, keys, and relationships in one view
Documented the company's core Visual Basic Library, the shared code behind their tooling, so a developer could trace how each routine was used
Implemented Pester and other tests for the core library so a change that broke shared code was caught instead of shipped
what I learned
A versionable schema doc beats a one-off diagram. The diagram shows the shape once; the doc, kept with the code and committed, stays true as the schema changes.
Legacy systems hide coupling. A field I flagged as unused often had a report or a batch job reading it somewhere else, so I checked call sites before recommending a drop.
Pester turns an untested Visual Basic library into something a team can change without fear. The first tests were the hardest, and they caught the silent breaks that motivated the work.
At 20 hours a week the scope has to stay tight. The value came from reducing unknowns: schema, diagram, and library documentation, each a durable artifact I would not be there to explain later.
An 8 month engagement shaped the effort. I aimed for artifacts that outlast the contract, because nobody else was going to write the docs.