Thomas Haighton, Research Software Engineer
I did my bachelors in Sonology (electronic music) at the Royal Conservatory in The Hague, this is where I had my first encounter with coding (Lisp/Supercollider/Max). After my bachelors I decided to follow a two year masters in Media Technology at Leiden University where I focused more on art installations / sound art. This is where I did my first project using natural language processing and data visualization; I created a few Max externals that could harvest Twitter data, perform sentiment analysis and display them on a map. During my masters life threw a curve ball and I decided to quit. I started an internship at Unilever’s Master Data Management Department in Rotterdam and follow Data Science (DS) and Machine Learning (ML) courses online in my spare time.
From 2016 to 2020 I have worked as a metadata specialist at the digitization department of the KB where I was in charge of the automatic quality assurance for digitized materials. This is where I learned Python to create my own quality checks – this is where my prior knowledge of DS really came in handy, because instead of just validating fields with specific patterns/default values or correct data structure I could create new checks based more on probability. I was asked to join internal research groups that dealt with enhancing OCR quality and automatic metadata. This was the perfect opportunity for me to get to work on some real world challenges using ML.
At the end of 2020 I got a temporary job as Research Software Engineer (as replacement for a colleague (my KB ML-sensei) who is on maternity leave) which gave me the opportunity to spend a lot more time on making cool stuff.