Building a learning culture — How we share our knowledge and grow

Today, I’m meeting with Geoffrey (Machine Learning Research Engineer) and Patryk (Software Engineer) who have taken the time out of their busy day to discuss their thoughts on Learning, Sharing and Growing. As I am relatively new to the business, I want to gain more of an insight into how the Research and Development department expand their knowledge and keep up to date with the latest trends.

Hello and welcome! So, what does learning and development look like at SO1?
Geoffrey: Personally, my background is in academia, I was a university lecturer in data science and therefore I have sound theoretical knowledge in Machine Learning (ML). When I came to SO1, the learning curve was quite high as I was getting exposure to real data and real clients. I improved my programming skills a lot. There had been a lot of added value to my previous background in academia.

Patryk: I feel like I’ve learned a lot from internal workshops. Recently, the Lead Data Engineer led workshops that focussed on the data pipeline. This was great because we don’t work with it every day as the Data Engineers do, we usually hear about it from the developer’s point of view. He explained the stages that the data pipeline goes through, how to onboard new clients, how to validate the data and even how to do the intake!

So Geoffrey, since your background was more theoretical, did you get enough support to make the transition during onboarding?
Geoffrey: Oh sure, I worked closely with other team members. They gave me the time to learn the applied skills that I needed in order to be successful. At the same time, I shared my theoretical knowledge to the team and we developed a great exchange. This still continues, even after all this time, as we are continuously helping each other out when someone is stuck.

Patryk: Yeah I agree, in the first month after I joined there were plenty of meetings explaining the business side of SO1. It was pretty effective, and I learned what we were doing, what the vision of the company was and some interesting parts of the system. When I came back to my laptop, my work was easier because I knew what it was for.

Patryk, you also mentioned internal workshops, what kind of internal training have you participated in?
Patryk: When I first arrived, the Software Architect here organized a session about the architecture and the set of goals we have for the new system that we’re building from scratch. He also set some time aside to show us Kubernetes (system for automating deployment, scaling, and management of containerized apps), which used to make me uneasy because of so many new terms. He sets time aside every sprint to run a workshop with us because we’re working in a real cluster deploying real applications.

Oh yeah, and we also have bi-weekly presentations where we learn new skills and see some live examples. One of the other software engineers in the team did a couple of presentations to show us what he was working on and facilitate a discussion because he had a problem with establishing the validity of the data. He showed us some examples and he’d figured out a materialized view and a complex query that wasn’t super obvious from the pull request.

Geoffrey: Ahh yes I remember that Kubernetes workshop. I hadn’t been exposed to it before so it was really useful to learn about it. Usually, I’m on the other side of the training. I’ve always been teaching. It must have been for ten or fifteen years now, so when I learn something new I like to explain the concept in a simple way. I spoke to the VP Engineering and we agreed that it would be a good idea for me to share this knowledge with the rest of the team. I ran ‘Machine Learning Fundamentals’ workshops to introduce the software engineers the most fundamental techniques of ML. They don’t necessarily need it in their everyday jobs but it’s part of general knowledge and understanding of what we are doing in the ML team.

I also presented some current research that I do on copula modeling and some research papers. The most recent talk was about BERT, which is a state of the art language representation model. We have a Journal Club every two weeks, mainly because we want to be up to date with research papers that are being published, and share our views on them.

Oh nice, do you have any other clubs like Journal Club?
Geoffrey: Yes, we also have the Bishop Reading Club that’s attended by members of the ML, Data Science and Software Engineering teams. We meet every week and solve a set of theoretical problems related to machine learning techniques and algorithms. It’s just for an hour a week, and each attendee is responsible for a task that they present to the rest of the team.

Finally, do you believe that Learning, Sharing, and Growing are important values here at SO1?
Patryk: It’s quite ad hoc, we like to learn as we go. We always figure something out and we talk about things over lunch or while having an after-work beer. We’re always growing and learning new skills.

Geoffrey: I really enjoy our social events. We have barbeques in summer and team breakfasts, it’s nice to have this light atmosphere at work and share our social lives with each other. We also have a 5k team run (Teamstaffel) coming up that SO1 will participate in, I’m not quite prepared yet though!

Haha, me neither Geoffrey, let’s go out later and train. The Office Manager has organized for us all to train together so we can all ‘share’ in the running pain!

Do you want to join a high-caliber team that will use your strengths and empowers you to fulfill your potential? View our open vacancies on our ‘Join Us’ page and we look forward to welcoming you to the team!


Retail technology news