Recruiting more skilled practitioners into the data and analytics industry is an urgent requirement, given the depth of demand from companies of all shapes and sizes. Ensuring those new employees come from diverse backgrounds should form part of that strategy. So to find not one, but two prime examples of the new talent now working in this sector is remarkable, to say the least.
Berenice Pila Dìez and Theresa Offwood-le Roux both hold PhDs, both moved to the UK to join this industry and can both now be found working as data scientists at Experian Data Labs, the seven-strong research and development unit of the data giant based at King’s Cross. Even more coincidentally, they both completed their transition from academia to the commercial world via the same London boot camp.
“For me, it was a simple decision to move from my PhD into the commercial world because my subject was very theoretical and was not helping the world. I wanted to apply my skills in a more practical realm,” Offwood-le Roux told DataIQ. “It is more satisfying to make products that can be put to use and create value. I would never go back to academia.”
A UK citizen originally from South Africa, she studied Financial Mathematics for her PhD, a subject that took her into the world of quantitative analytics in a bank. “I liked the maths and the problem solving, but not that financial trading industry, so I started to look at where else I could apply myself,” she explained. Via a move to London and that boot camp, she first took a job as a data scientist with a start-up before joining Experian in November last year.
For Pila Dìez, a degree and Masters in Physics in Spain led to a Phd in Astrophysics in The Netherlands. “I realised that I enjoyed the subject, but I wanted to do something faster-paced and which would have tangible impact,” she said. After moving to London and transitioning through boot camp, she worked for one of the big four management consultancies before joining Experian in 2016.
Experian Data Labs was only set up two years ago with a focus on helping clients to solve complex problems using their own and the firm’s data. “There is a lot of data out there which they don’t know what to do with, so we help them,” said Offwood-le Roux.
Examples of the projects which the team work on range from classifying and scoring transaction types into meaningful groups to help both business and consumer customers towards better financial health, or improving the identification of fraudulent activities to reduce false-positives and thereby reduce the workload for fraud teams. Both involve combing through high volumes of data for meaningful patterns which might be the root causes of a business problem.
“It has not been much of a change for me because my PhD was very data-centric,” Pila Dìez pointed out, “and the techniques are very similar. The biggest difference comes from communicating with clients and managers - that has required a level of change and thinking outside the box. You need to express things differently.”
Offwood-le Roux agreed, but laughingly noted that the effort to translate between the two groups comes more from the data scientists than the executives. “Thinking about questions in a commercial way is a challenge,” she said. Equally, she argued that many organisations have rushed to hire data scientists as the new must-have skills set, when they should start by solving more basic problems: “Don’t just go straight to machine learning.”
Just as important, if an organisation does adopt data science, is how it is managed. “You have got to keep the two environments - production and R&D - separate as they are very different with their own approaches and cultures,” noted Pila Dìez. As data science matures, some of its practices (and outputs) will migrate into the day-to-day business, but the need to innovate will remain.
That will create the conditions for continuing recruitment and, with that, the need to ensure the widest pool of candidates is tapped into. Ensuring that female students form part of that future continues to be an issue and one that both Offwood-le Roux and Pila Dìez are keen to tackle. While neither has experienced the “brogrammer” culture identified by Sarah Noyes at Speak with a Geek - Offwood-le Roux said, “that is more dominant in the US than here” - they have worked in male-dominated environments.
“We need to make STEM more relevant for girls - boys get more involved through gaming and coding,” said Offwood-le Roux. Her own desire to change the world has led her to volunteer for the data-for-good network DataKind UK, while Pila Dìez has already taken part in several STEMnet events aimed at fostering an interest among children and teenagers. But, not least on International Women’s Day, it is their bold decisions to change careers and countries in order to pursue a love of data science that should serve as role models.