Q&A with an Aire Data Scientist — Chris Howlin
Chris Howlin (@cristohowlo) has been a member of Aire’s Data Science team since February 2018. Here’s a quick snapshot of his work, and life beyond.
- Coffee count so far today (10:36)
☕️☕️/️☕️☕️☕️☕️☕
Aire-time
2. What’s been your biggest accomplishment since becoming a Data Scientist at Aire?
Actually becoming a Data Scientist! I joined Aire as a Software Engineer, and moving into a Data Science role was the culmination of a couple of years of learning and planning. Also attending the UK CCTA awards for the work we have been doing with Toyota Financial Services was great fun and good recognition of the work we have been doing over the past year.
3. What are you most looking forward to getting your teeth stuck into next at Aire?
Most recently, I have been working with a colleague to introduce Bayesian methods into one of our models. There’s a lot of power in the specification of prior knowledge and the direct quantification of uncertainty. I am keen to see how we can develop our learnings into wider practice.
4. What uncomfortable forest* are you glad to have taken a walk in since arriving at Aire?
- At Aire, we’re not afraid to take the hard path. Even if it takes more time or creates more pain. We want to ensure we do it the right way.
A couple of months ago I had to stand in for our CTO, Tim, at the last minute to be on a panel hosted by the Digital Banking Club at the Law Society about the use of AI in Motor Finance. This was the first time that I had been on a panel in front of a group like this, so it was quite nerve-wracking.
Fortunately, the team at Aire was great in helping me out with the preparation, and the event went very smoothly. If you’re interested in some undiluted opinion (and seeing my first time on camera 😬), there’s a video with panellist interviews.
5. What do you think the biggest change to the Data Science team has been in the last few months?
After staying the same size for a while, the team has grown a lot in the last couple of months. It’s a great sign of growth but we have had to re-think a lot of our ways of working and take on all the new thinking.
Downtime
6. How do you stay up-to-date and keep your skills sharp?
I help organise the London Data Science Journal Club Meetup, which gets together monthly. It’s like a book club for the latest academic papers, the content can be pretty demanding but we cover a wide range of topics! For example, in recent sessions, we have covered the estimation of Shapley values for explaining model factors and transfer learning for text classification.
7. What are you learning right now?
Reading The Book of Why (Pearl and Mackenzie, 2018) earlier this year, I have gone through a huge shift in my mental model of Data Science. Causal inference is an under-discussed area of research outside of epidemiology and medical research, and I see huge potential for it in the area of credit scoring…
8. What’s been your favourite conference experience?
PyData London is always a great experience: a big shout out to the volunteers who put in so much effort to make it a success. The quality of talks is incredibly high; it’s like an annual coming together of all the interesting people you meet during the year.
9. What’s the one conference you’d like to go to but haven’t been to yet?
It would be great to go to one of the more academic conferences. One that stands out in particular is the recent ACM FAT* conference, a multi-disciplinary event that centres on the research around fair, accountable and transparent directions in machine learning.
10. What’s the podcasts/ blog/ book you swear by?
More than podcasts or blogs, I find Twitter keeps me up-to-date with what’s new. Follow people who publish research you find interesting, see who they follow, and pretty soon you will have a high-quality stream of the latest discussions in your areas of interest. For starter recommendations, you can use the various unofficial ArXiv feeds to notify you of the latest updates to the different categories. I find @StatMLPapers has a good hit rate. Also Judea Pearl (@yudapearl) tweets regularly about his work in causal inference.
11. Any personal projects outside of work?
Nothing right now, but I am actively looking for some open-source work to attach myself to. If people are looking for an opportunity to start giving back, I would recommend the wonderful Python Sprints group. They are great at supporting people who have never made a commit, and helping structure improvements to packages which we all use and take for granted!
