BAI Interview Series: Yves Mulkers on Data Science Skills


We recently interviewed Yves Mulkers, founder of Yves is a Business Intelligence and Data Architecture consultant with twenty years of experience in the field.  We wanted to question Yves on which entry-level skills are  required of management in manufacturing or service organizations today particularly in data science and analytics. 

Tell us about your work and how you ended up in your line of work.
Back in 1996, I began a long and often passionate relationship with I.T. I was a kid when the Commodore 64 came out. It was my first home computer I found it amazing.  I taught myself at the time basic programming. We learned some of the basics of informatics at school, but my interest was already piqued by all the new technology at the time and the things that you could do with it. But I graduated as a chemist. I’ve been working in various roles for about six years now on operational environments.
I was so passionate about computers and what was going on at the time. I kept up with the technology throughout the 60s, 70s... always revising my knowledge with each evolution of technology.  At the same time, I was deejaying so I had a very big vinyl collection. And I had a hard time getting the collection organised. In my search for a database that could store all the music I was steered towards a product called FoxPro which I used to build my own music catalogue. And that’s how I taught myself to program using Foxbase/Visual Foxpro. I felt more confident in my ability to code, and it opened opportunities for FoxPro programming jobs. It was the first opportunity to specialize in IT. But I knew at the time I had never done analytics of any kind and that I needed to acquire those skills. 
So that was my challenge. On one hand, I wanted to gain new technical skills.  I thought at that time the more I knew about programming the more valuable a specialist I would become. On the other hand, I learned that having technical skills were not the only thing, I needed to know I also needed to understand the basic concepts on data management. So, I came across SQL server databases but I was not too interested in the technical coding side of it. I was more interested in understanding where we go with IT than what was going on under the hood. 
I came across Business Intelligence while working on a project. I thought it was a good balance between IT and business.  From a technical perspective, I had always been interested in what a business does, how they do it, how it can be optimised. I wanted to contribute to businesses using my technical knowledge and Business Intelligence.
Over the years, I have always challenged myself to keep learning so in addition to my technical skill set and analytics, I learned conceptualising, communication, visualising, which are what I would call soft skills.  So that’s the road I have taken to get me where I am today and to what I do in my day to day job. In addition to that, I am also interested in social media in aggregating and then analysing the data. So now it’s all these experiences kind of coming together into a coherent whole. That’s been the journey I’ve taken for the last 20-25 years.
Which technical skills, in referring to data science and management; do you feel a management candidate should have? 
I don’t think having a deep technical background is necessary for a managerial candidate. It’ is important to have   the technical know-how or the capabilities though even if they may have never implement a script.  These are people that will probably start from the managerial level and should have the from the conceptual view of a business intelligence project. A manger would take the role of project lead with a more top down overview of a project and should have with a very good understanding of the business. They would also have a good grasp of the technical problems to the point that they are capable of the challenges without doing the technical work themselves. They should at least be able to manipulate the data: if they are in a data-driven business environment they should be able to aggregate raw data and identify the outliers or what data is missing. They should be data savvy- able to look critically at the pertinence of data even when it’s not their own. 
People at the managerial level should build the vision and the strategy on how to get to a solution and be able to communicate the bolts and nuts of analytics. The managerial level has less to do with the low-level details than proof of concept and transforming data into impactful decisions. 
Cognitive skills refer to an understanding of how people interpret the data they see, as well as how they use data to incite action. Which specific skills do you feel an entry-level manager should have today?
Cognitive recognition, or put more simply, being data savvy, is when you get the feeling for your data. You know if it is pertinent to do aggregate SQL query where you are joining data together you can immediately tell it the outcome result is correct or not.    Maybe you didn’t query the join yourself but can point out to the team that it’s double the amount that doesn’t seem right. Cognitive recognition also refers to the ability to recognize errors in data.  Some people are who are number savvy are very good at pattern recognition. 
Other key cognitive skills focus on the ability to communicate. So, for example, if you build an overview you should be able to explain what it is in it and what it means for example, why do we have deviations and you can, really in human language, discuss with people about the data, not in a way that just describes the query but in a language that people from sales or marketing can understand.
Trade skills refer to an understanding of the business logic, organisation, and monetisation of company resources in each market. Which specific skills do you feel an entry-level manager should have today?
For trade skills, I think how fast one understands the business is very important. There are two levels of business understanding in my mind. There is a global understanding of how an organisation works. By this I mean   sales, finance, marketing operational, external customers, and internal customers. There is also industry specific knowledge. There are some skills that you can transfer from one company to another. It’s very useful as well, I believe, if they have worked across various industries so that they can carry across the lessons learned and   conceptualise them to transpose those skills.
Trade skills can be demonstrated in exploring how fast the potential manager understands what the business is about and, understand what is important to the business. For example, businesses primary focus is on the markets. When they feel they have captured the markets they then move on to optimisation of the revenue. That’s one example of an approach that some businesses have.
Another example is a small business, let’s take an electrician who has been walking long hours and long days but he is not making a profit. He is not considering the profit margins when he calculates their prices and costs. My approach is to ask a business what their vision is. Where do they want to be? What numbers do they need to improve? So, I like to talk to the business leadership and ask them, what is your vision?
Lets’ take another example of a big business, an insurance company.  The salesman suggests that the cost is not a problem. That tells me that the company has a large budget and I inquire about their revenue.  I’m trying to understand why they are interested; or not, in costs, revenue, or profit margins because if I understand those numbers then I can help them optimize their business When a business says,we want to do more’, thedo more’ must be understood. Is it profits, margins, or costs? Trade skills are needed to understand what thedo more’. Having a good grasp on trade skills allows to know which are the low hanging fruit can be readily harvested and where one can get the best return on investment.
In your opinion, how should organizations or HR agencies, evaluate these skills before and during employment? 
Well traditionally, HR tests your soft skills, how your communicative, what kind of person are you, do you fit within the team and so on and so forth. If the HR specialists have been working in that domain already for a longer time, for instance data recruitment, they will understand the general terminologies in as much as they need to know. However, they struggle with confirming how well a candidate will fit into analytical jobs. 
It is possible to evaluate capability and skills in an interview. I have had various interviews where they were good at doing exactly that. The best example I can give is with a recent interview I did where I was tested on an SQL script. So, they showed me the query and asked me to talk through it for instance, what would be the result, what do you think is happening. They also asked questions like, if you need to add data to our sandbox how would you do it, would you use Excel or write a script. I had to explain my preferred answer to the question and the reasoning behind it. There wasn’t a right or wrong answer and the interviewer just wanted to assess how fast I would arrive to a solution and my reasoning behind it.  I was also assessed on how well do I would fit in with the team.
In this example, the project manager led the interview- he was capable of putting together the infrastructure and leading the team because he was so proficient in all levels. I believe it was because of his proficiency that he can do such a pertinent interview.  There are poor experiences I have had during the recruitment process in talking to recruitment agencies.  For instance, a recruiter asking me to explain dimensional modelling only to inquire a few minutes later if I was familiar with star schemas(a star schema is an example of dimensional modelling).   It appeared in they were reading a script-  asking me questions about key words on the job profile. I cut the interview short because I didn’t feel confident that they were able to evaluate my skills. I think a good selection process should integrate good use scenarios and practical hands-on assessment which would guide the interviewer in his or her evaluation.
Yves Mulkers is a Business Intelligence & Data Architect consultant and a Social media Influencer His LinkedIn profile can be viewed at can follow him on Twitter: @YvesMulkers