“What do you do?”
It’s a question often asked, but one that I sometimes find hard to answer, especially when everyone else has preconceived ideas about what different jobs involve and the kinds of people who do them.
I’m not a software developer – writing code is an important part of my job, but software isn’t primarily what I do. I didn’t learn to code until I got to university and realised I enjoyed it as a way to build stuff. I don’t enjoy coding purely for the sake of coding, and I get lost in discussions of the more theoretical aspects of computer science.
I’m not a scientist – I do plenty of experiments, but I work in the Engineering department of a university where the main driving question is ‘what works?‘ instead of ‘how does it work?‘. Science fundamentally tries to find out how the world works, while Engineering tries to build on our scientific knowledge to build stuff that’s useful. I’ve always been more interested in practical solutions than theoretical ones; so while I enjoy science, discovering the ‘how‘ is not my main motivation.
I’m not an engineer – People think that Engineers wear hard hats and spend their days outside building things. In contrast, software is intangible, just a bunch of 1s and 0s somewhere in computer memory. But actually, Engineering know-how is just as useful for building things out of software as it is for building things out of metal and concrete.
In reality, I tread a fuzzy line between industry and academia, part-engineer, part-developer, part-scientist. I work on improving speech recognition technology, and spend my time researching to find out what works to improve our current state-of-the-art. My job splits roughly into five main responsibilities:
- Reading literature to keep up with recent developments and coming up with new ideas
- Writing code to implement the new ideas
- Writing more lightweight scripts that run a bunch of programs, collate results, analyse data etc.
- Planning and running experiments, then analysing the results to see whether the new ideas are any better than the old
- Reporting and paper writing
In the end, I’ve given up worrying about trying to fit into one box, even if that means I don’t have a snappy answer when I’m asked “So, what exactly do you do?“