It may be the engineer inside me, but I find optimizing systems to be highly challenging and rewarding. I enjoyed the time spent analysing and refactoring SQL code and more recently have thoroughly enjoyed refactoring some of our existing applications to run more quickly.

The great thing about optimizing computer code is that you genuinely know if you’ve succeeded or not. You have metrics from before and after your code changes to prove they were effective. It also makes you think about how things are working in the application at a much lower level. It feels pretty good that by you spending a couple of hours on a problem you could be potentially saving days, weeks or months of run time when applied to a number of users.

In this article the author details how he went about improving the speed of a facial recognition implementation by 25 times. That’s a massive improvement and I found the process to be pretty nicely indicative of how a real programmer thinks sometimes. If you consider that this process perhaps took him a day of work, and returned some software that potentially requires 25 time less equipment to run, it just goes to show how much difference one person really can make!