Deirdre Dempsey has decades of experience dealing with data analytics, data science stemming from a strong foundation in mathematics.

Deirdre Dempsey leads Three Ireland's advanced analytics team and data science development. Having worked for more than 20 years with data, she is a keen advocate of data-driven decision-making and harnessing data's potential for business. She comes from a grounding in mathematics, with a degree in mathematics and mathematical physics and a master's in mathematics from Maynooth University.

"While the emphasis of my role has moved over the years away from being the data practitioner, I do still love to work with data whenever I can," she said.

Dempsey talks more about her role - where "no two days are the same" - and the benefits of a mathematics background below.

You've been working with data for decades. Have you seen dramatic changes in that time in terms of how data is used and valued by businesses?

Has it been that long? I wouldn't attempt to quantify the change that I've seen but I can testify to the pace, particularly in recent years. The length of time for innovation to be implemented has become drastically shorter and cheaper. I remember being very happy to gain a Zip drive or CD-ROMs with the analysis software, as opposed to numerous floppy disks, and now we have the phenomenon that is big data, open source software, and easily downloadable tools.

As to how data is used and valued by businesses, I can summarise this very simply as the move from 'I know my customers, I don't need data to tell me about them' to actively seeking data to support and drive decision-making, and onwards to the expectation of proactive innovative solutions through data. This change has been driven by consumer behavior, by industry knowledge, by awareness of the benefits that data can deliver and also by the sheer volume and accessibility of data in recent years.

When it comes to communicating data science and analytics to other stakeholders in your organization, has that become easier as data literacy improves, or more complicated as technology becomes more complex?

It has definitely become easier, but this is twofold. With the growth that we've seen in recent years and the acceptance of data as an important element in decision-making, people are hungry to understand what data and analytics can do for them and their plans and business cases. There is also a growing awareness through the ever-increasing number of conferences and meet-ups. Also, there's now a trust for data and an acceptance that there's a science that the experts understand.

I've worked with data in business for a long time and I've had to develop the skill to communicate data science in plain English. I've achieved this by sometimes getting it wrong and learning from it. As a result, I'm working to constantly improve how we tell the data story.

Tell us about your background in mathematics and how that led you to work with data.

I've always had a strong academic background and particularly for the sciences. I was quite young seeking out my third-level choices and opted for mathematics because I enjoyed it, I was good at it and I had no idea how else to choose!

This, thankfully, worked out for me and I pursued a degree in mathematics and mathematical physics followed by a master's in mathematics. I taught maths for a while (which I loved), dabbled in auditing and found myself in the data science world quite by chance. I had a colleague from tutoring who recommended a role to me that I might not have spotted. This led me very quickly to a 'this is what I'm meant to do' moment and I haven't looked back since. I found my niche in working with data and teams of data specialists.

Do you think it's commonly understood that skills in mathematics underpin data science and analytics?

I don't believe that it is. Mathematics lends itself more to the traditional statistical sciences rather than data science, which may be seen as being more in the engineering and computer science realm - albeit there are shared core competencies and capabilities, particularly statistical knowledge and concepts that underpin a lot of what we do. In fact, the most common assumption coming out of my primary degree, although decades ago, was that I must be going on to either teaching or auditing because what else could you do with a maths degree?

Could this awareness make mathematics study more attractive?

General awareness of the sector overall is needed both in terms of the role and what it actually entails and the routes into it. However, we need to start with those already studying mathematics or statistics. I have met many graduates who've asked me the question: what can I do with a mathematics degree? And who've been both amazed and perplexed when I explain data science to them and the realization that this is a career path within their reach.

Maths and data science are two sectors known to be male-dominated. Have you noticed this yourself and do you see that pipeline changing in future?

Yes and no. I have been privileged to work with some very talented female data scientists over the years. However, what I've noticed is that, while there may be fewer females in the data science sector, we need to go further back before career choice to see the genesis of this.

In my experience in recruiting and developing teams - and indeed borne out by research - females are less likely to pursue the STEM qualifications that lead into the sector. Therefore, the change needs to happen earlier than the entry point to the sector, and even the third level. I do see a very positive change here with the programmes that are being put in place nationally and internationally, and I look forward to seeing that flow through to the sector.

Who are your own heroes in maths and data science?

This is probably the most difficult question. How to pick from so many talented people! Fermat, for the conundrum he posed, Schrodinger for his logic, Boole for Boolean algebra and its subsequent evolution, Thomas Davenport and DJ Patil for their work in advancing data science, Andrew Ng for making quality learning so accessible and, of course, Doug Cutting for giving us the means to access and analyse vast amounts of data.

But also my team for delivering, no matter what I ask of them, from transforming processes to figuring out how to revive dead servers and code, to constantly looking for ways to innovate, automate and use data better.