Always asking why as she mines data
Megan Yates is the founder and chief scientist of data-led modelling and analytics company Ixio Analytics. She tells Margaret Harris that she enjoys being able to get to grips with and solve problems
You are both a data scientist and an evolutionary biologist — tell me what these entail.
A data scientist processes large amounts of data — from various sources and in diverse forms — to discover insights that drive future actions. The focus of the job is on improving business outcomes. I love the investigative, analytical process and seeing real datadriven value for our clients.
The field requires excellent problem-solving skills, using a variety of mathematical and statistical tools and techniques, to make sense of complex business data. Biological data, with its often large and messy datasets, isn’t too different from business data.
An evolutionary biologist studies the genetic, ecological, geological and environmental factors and processes that produced the patterns we see in nature. The field is diverse, and my research focused on form and function in proteas and restios (Cape reeds). The questions asked in evolutionary biology are often simple — for example, what is the advantage of small leaves? — but generate answers that invariably lead to more questions, requiring a multidisciplinary approach.
What do you do at work?
Data science usually starts with a problem such as “Most of my customers don’t pay me”. The first thing we do is to gather as much information and data as we possibly can. Then we play detective and try to understand the patterns hidden in the data. We use many different tools and techniques to do this.
That allows us to do something really cool: predict which customers won’t pay. Knowing which customers might not pay us at the end of the month allows us to do something about it, and that’s very valuable.
What do you most enjoy about your work?
I find the problem-solving part enormously rewarding. There is huge fulfilment in understanding a problem and then working to solve it. This involves setting up different experiments and trying various methods to find the best solution.
What characteristics do you need to do your job?
The most important trait for this job is curiosity. This, combined with good analytical and problem-solving skills, means continuously figuring out how to do things better, challenging norms and assumptions and always asking: “Why?”
Another vital characteristic is to be a good communicator.
Being mathematically, statistically and technically skilled is a prerequisite for the job, but the ability to communicate solutions and results to clients separates the best data scientists from the others.
What would you do if you had to choose another career?
I’m so focused on data it would be hard to choose a career that doesn’t involve data in some way. If, for some reason, the field of data science was no longer an option, I’d be interested in a career in military intelligence. This field involves enormous data challenges like formats, timing, space and importance.
What did you want to be when you were a child?
I desperately wanted to be a forensic scientist. I loved the investigative process and using several disciplines within science to solve mysteries.
I worked in a home interior shop while studying. The ladies running the shop worked incredibly hard and taught me that dedication and perseverance pay off.