Uncovering Hidden Talent: Interview with a Lead Data Scientist: Ibrahim Shore

Hi Ibrahim, great to speak with you. Could you introduce yourself for those that don’t know you?

Thank you for this pleasure Angus, it is much appreciated. Of course, I work for a full-service agency - VaynerMedia as the Lead Data Scientist and have been here now for just over 9+ months and the journey has been incredible.

Great, thanks for that. Well let’s start here: for those that don’t know, what do data scientists do?

Sure, in simple terms the job of a Data Scientist is one where we seek to understand the challenges and objectives of the business. In so doing we can go away to devise a plan for how we will predict or forecast the future based on the data we have today using a range of statistical methods.

That sounds like a useful function for a company. So what was your journey, and why did you choose Data Science?

This is a funny one because I never chose data science as a path initially, I set out to be an Economist, but the route was a difficult one. In my first job after University I started off as a Claims Handler for an insurance company and whilst I was there a secondment position came up for a Data Scientist. So I scanned through the job spec and the role sounded fascinating so I applied. It started off with a secondment position but I was just grateful to be granted the opportunity. I cannot tell you the hours I invested in a few courses I found on Udemy but it was certainly worth it.

What advice would you give someone who is looking to get into Data Science right now?

Oh boy, this is a good question. With there being a huge number of people wanting to get into the field I’ll give this one piece of advice, something I like to call your differentiator. Build your own GitHub repositories. It's like an Artist wanting to land an opportunity without a portfolio, it just doesn't happen. I also wished somebody had told me about the PEP8 standard of writing code earlier, this for me is a huge win if someone masters it. Let me end my answer with this: whatever you do, never sell yourself short by doing a role that is completely different to data science, all because you want to eventually get there. I made this mistake and it took me a while to get back.

What are some of the mistakes people make when first trying to get into Data Science/Analytics?

I find that people often spend too much time in theory and when it comes to its application, they are lost. The best way to solidify learning is by practising. One must be very good at applying what has been learnt from the point of theory. People give up or become stuck and demotivated because they cannot see a link between what they are learning and what is done in the real world. Another is being unprepared to discuss previous projects done. So review and practice describing previous projects you've done, whether that is in previous roles or ones you have done in your own time. Know it through and through.

When you are looking to recruit a junior-mid Data Scientist, what do you look for?

Good question, on this point I would like to highlight a few things. The first is having the right attitude. One might ask what does this really mean? Well, to me this means being grateful, showing empathy, gladly taking on responsibility, respect for others just to name a few. The next is having good technical skills whether that is being proficient in Python, SQL, R. But the other I would say is being able to communicate insights to a non-technical audience. All of the above differentiates a candidate.

What is a Data Science interview like and how would you suggest candidates prepare for them?

A good way to prepare for these is first to understand the job spec, this is often a clue to what they will be testing for during the interview process. Outside of that, it is to brush up on foundational topics when it comes to statistics and machine learning techniques as well as understanding Python, this could be on how to write loops, functions etc. A good way also to prepare for these is to take on some Kaggle datasets and play around with the data. 

You mentioned a few coding languages a couple answers back, so let’s talk about them. What are the important languages to learn to progress your career as a Data Scientist/Analyst?

The first that comes to people's minds is Python and I'll have to agree simply because of its sheer power and the community support is immense! Another that is often overlooked is SQL. Though my next suggestion is not a language, I would strongly encourage both people who aspire to get into this space and those already in the boat to learn cloud computing, whether that is Microsoft Azure, AWS or Google Cloud Platform. I cannot stress this one enough really.

What would you say to people who say they can’t do data science because ‘they aren’t good at maths’?

Then I have some great news! If you are good at programming, have strong intuition, a good grasp of understanding business challenges then you are not far off starting in a role in data science. A huge part of data science is statistics and probability and will take effort and persistence to develop, not going to kid anyone about this. Back to maths, if this is not something you feel you are not yet good at, it just might be that you have not found the right avenue to learn this. Find the right course or a person who delivers it well. I came across this article sometime ago and found it to be incredibly insightful. I hope you have been encouraged by this.

What is the latest in Data Science, and what’s around the corner?

There is now talk of a fairly new programming language replacing Python eventually called Nim, whether that happens is anyone's guess. Also, as we get to the age where physical servers are slowly becoming redundant the world is moving fast to cloud computing. To add, there is so much new technology in this area such as Docker and what an incredible tool it is.

Thank you for sitting down with me, Ibrahim. Any final thoughts to share?

Just grateful for this opportunity to share my experiences with people, especially because I consider myself to be a self-taught data scientist. And I genuinely hope my experiences shared helps someone to get to where they want to be.