My brief moment in HR

And now for something completely different.

After spending a lot of my career in physics and mathematics, I thought why not try my hand at something else. So I decided to spend 10 months working for an HR start-up leading the software development and data science team. Yes, HR as in Human Resources! The aim was to leverage machine learning (mainly supervised learning) to reduce absenteeism and turnover rates at large companies and then put this all in a nice web based application. Long story short it was very interesting and a great team to be part of but 10 months was long enough for me in HR.

Besides getting frequently asked if I was doing this for a joke, and having to get used to the pseudo science of HR analytics (a poor grasp of statistics for one), there were some interesting problems to solve and the tech stack was really nice. It was Django + (DRF) for the backend and React frontend, using Sckit Learn and Tensorflow for machine learning. We developed a lot during a relatively short time (and switched between many projects and applications in this time) and I am quite proud of what we achieved - a nice algorithm for predicting absenteeism using an interesting Bayesian approach … but I can’t tell you much more than that. Also, I got to make some fairly sleak visualisations using D3 and wrote some nice python code trying to map a relational database to a graph model.

Ten months of HR was enough for me to tolerate, and the joke is now up, and I am now safely back in the realm of physics. (I actually wrote a small piece a while back on some experiences of my first HR conference compared to typical physics conferences - the differences were stark, read about it here.)

Anyway I also gave a 20 minute talk with Al Ademsen about new techniques we can apply in HR from other domains such as marketing, politics and fusion research. See the full video here.

Written by Thomas Stainer who likes to develop software for applications mainly in maths and physics, but also to solve everyday problems. Check out my GitHub page here.