Celebrating some unsung wins

When you are the only programmer on a project, you sometimes get an unsung win, something which you are proud of, but that no one really understands so didn't notice fully 🥲. Often it's little things, like being able to add a new feature really easily after 'past me' did well on the software architecture. I had a few bigger ones recently, and so here's a little post to celebrate.

  1. I built a two-way contract for linking remote datasets, creating a redundancy which proved to be useful when one side had intermittent failures when storing an id. This saved a lot of time which would have had otherwise required messily manually linking datasets.
  2. An expensive pen-test was performed on a large project, and the app side of things (my responsibility) passed without any needed modification. One thing I think helped is a principle of ignoring the siren call of bleeding edge stacks, instead preferring mature, stable and whence battle-tested technologies.
  3. I used a combination of generated SQL views and python to automatically pre-process tables of raw data for 40+ experiments saving countless hours of manual dataset merging and tidying which would probably have been undertaken by the future analysts of the data.