11 April 2018
Jonny Brooks-Bartlett - This interview was conducted by Erica Freedman,
Content Marketing and Client Service Specialist at SwitchUp.
Jonny began his career in Mathematical Biology, but was fascinated by the complexity and skill-level of data scientists. Finding S2DS, Jonny was able to transition into a Data Science career. Working in consumer experience, he curates customer facing tools using a data-driven model. Read more about his journey to data science and how he was able to pursue his passion of building things people actually use.
Your educational background is in mathematical biology. What ignited the switch to Data Science?
During my DPhil (that’s the name for a PhD at Oxford University) I began studying a certain class of algorithms called Kalman filters. It required me to think very differently about the problem I was trying to solve. I was fascinated by the fact that I could frame a problem differently and still solve it regardless of the field I was applying it to. I was more interested in the solution to the problem than the biological problem itself. When I realised this I started looking into careers that had a focus on algorithms. Data science came up as the number one field for this and from then on I knew it was the career that I wanted to do.
Do you feel Data Science and Mathematical Biology are related? If so, how?
Mathematical Biology and Data science are related to the fact that data science is such a broad term that it would likely encompass mathematical biology in an industry setting. By that, I mean that the type of work I did during my DPhil could easily be called data science if I was doing that work in industry. I was using machine learning algorithms to solve problems. So the skills I developed during the DPhil are similar to the ones I’m using in my current data science role. As an aside, I’ve realised that many of the terms thrown around in industry (e.g. artificial intelligence, machine learning, data science etc.) are blanket terms for a broad range of things. In contrast, the names used in academia are very specific. So you could potentially be doing the same thing in academia and industry but have different names for them depending on where you work.
From the outside looking in, the data science field seems to be full of incredibly smart people that are infinitely more knowledgeable about a wealth of things (technology, algorithms, visualisation etc.) than I am. This meant I didn’t believe that I possessed enough of the required skills to be a data scientist whilst I was still in academia. So I figured that a data science bootcamp would bridge that gap. I found a couple of data science bootcamps in London and S2DS was one of them. I was really attracted to the fact that you had to work in teams on S2DS projects. The cherry on the top was attending a webinar that was hosted by Kim Nilsson. She sold the program to me in that webinar so I applied and was lucky enough to get a place in the August 2016 S2DS London cohort.
You currently work as a Data Scientist for Deliveroo. What does this title mean and what does a normal day at work look like for you?
I currently work as a Data scientist at Deliveroo and was at News UK before that. Currently, I am in the consumer experience team which looks after the customer’s experience from the moment they place an order. This means that I’ll write algorithms that will hopefully improve the user’s experience while they are waiting for their order to arrive.
A normal day for me varies quite a bit but it will usually start off at the gym! I love to go before work. It gets me ready for the day and I feel good when I start work at about 9am. I’ll have a stand up meeting with the consumer experience team in the morning to find out what everyone in the team is working on that day and whether anything will impact me. After that, we might have another standup meeting with the other data scientists to find out what they are doing and whether we can help with suggestions for their work. Afterwards, I’ll get down to work, analysing or cleaning data or even writing a machine learning model if I’m at that stage of a project. Around midday I’ll have lunch with some members of the team, whether that’s consumer experience or data science. In the afternoon I have a couple hours to get on with some more work but depending on the day we may have a technology department gathering to find out what’s going on with other tech teams. Or I might have a study group where the data science team gets together to discuss a topic that they’ve decided to study. At the moment we’re learning about experiment design (relating to A/B testing and its variants). Sometimes, I’ll have meetings with other members of the team or stakeholders to get some guidance on the project/model that I’m working on from a business perspective. I’ll typically leave the office sometime between 5:30pm-6pm. Often I used the commute time on my journey to and from work to read about data science tools/algorithms or literature related to the study group topic. So my data science learning doesn’t usually stop when I leave the office but typically it lasts until I’m off the train.
How is Data Science different in the sphere of news and public information?
I often find that public information about data science is overhyped and sensationalised. I do read articles about data science in newspapers if they are there but I take it with a massive pinch of salt. They also mainly write about the stuff that huge companies like Amazon, Google and Facebook are doing which isn’t necessarily that representative of where most companies are at in their data capability. Often you hear that some crazy machine learning algorithm has disrupted an industry or that artificial intelligence is going to take all of our jobs and kill us. As usual, the truth is somewhere in between.
Have you faced any challenges trying to become a Data Scientist?
Yes. The challenges that I faced trying to get into data science is typical of any University graduate trying to get their first job; lack of commercial experience. So that’s not any different to many other graduate-level jobs.
The main challenge I’ve faced being a data scientist is people not understanding what my job is. They seem to hear data and think that I can do magic with it i.e. things that are beyond my knowledge and current capability. This is probably reminiscent of the role of the ‘webmaster’ in the 90’s. Back then the webmaster was considered a unicorn that knew everything about websites, how to design and build them, maintain the servers, setup email and much more. Now you barely hear the term used in job roles anymore. You have separate front-end and back-end web developers, UX and content designers etc. I think data science will go this way when industries catch on to the fact that the current role is very broad and new tools and technologies will only exacerbate this problem.
Has S2DS helped you to get a job in your field? If so, how?
Yes definitely. The obvious thing is that it gave me experience on a commercial project that I could put on my CV and talk about in interviews. Additionally, it’s provided me with a network of people that have helped me throughout the last 18 months. Whether it be help with particular projects or getting me in touch with companies and recruiters. With all that said, the reason I got my first data science job at News UK was that I was invited to give a talk about my S2DS project at a data science meetup organised by Pivigo. My eventual employer was in the audience and approached me after the talk to consider a data science position at the company. The rest is history.
Where do you see your career heading in the next 5-10 years?
It’s hard to say for certain because things change so quickly. I imagine that I would be in a managerial type position. Over the last year and a bit, I’ve started to really appreciate the importance of working in a happy and cohesive team. So I’m trying to take initiatives to encourage a positive team environment on top of my individual contributor duties. I also love communicating with people too. I think that these type of skills will lead me towards a more managerial type route. But I do think it’ll be quite hard to take a step back from doing the technical work.
What makes you most passionate about the world of Data Science and Mathematical Modelling?
Two things. I’ve always loved being able to explain the world around us with mathematical models. It’s the reason why I loved applied maths so much. For me, the natural extension to that is predicting and classifying things with complex machine learning models. So I love it. Secondly, I love building things that other people use. When someone says,“This is amazing, it’s been so useful”, it fills me with absolute joy. So that’s the other thing that drives me.
If you could go back and give yourself one piece of advice before pursuing this track, what would it be?
Trust in yourself. It sounds so cliche and wishy-washy but I think it would have helped me. When trying to do anything new, people tend to doubt their ability to do it and second guess themselves. I’m the same. I often thought that I may not be good enough to get into the field and I definitely didn’t think that I would be as effective or as desirable as an employee as I am at this stage of my career.
With that said, I’m happy with where I am and if I’m happy with where I am then there’s no reason why I would change the path I took.
If you found this Use Case interesting, why not review the other Use Cases in our archive.