What is your academic background?
I originally completed a Bachelor of Science degree in Geography and Topographic Science at Swansea University followed by a Masters in Photogrammetry and Remote Sensing at University College London. These degrees gave me a great basis in mathematics, data analysis and the fundamentals of image processing. These were completed several years ago and the use of neural networks for image classification was talked about as a future tool.
Have you completed any other training in data science? (Up-skilling, MOOCS, short courses etc)
In 2015 I went back to University and studied a Graduate Diploma in Business Analytics through Deakin. I completed this remotely whilst balancing a full-time job, a young family and house renovations, it was a tough slog, but the change of career was well worth the effort. Studying for a second time gives a different perspective to the motivation for it.
If you pivoted into a data science role from another area, how did you go about this and what advice would you give others looking to do the same?
Between my masters and my move to data science I worked in the geospatial industry, primarily as a geospatial consultant for several global consultancies. Following a redundancy, I took the time to assess and reflect on the elements of my job I really enjoyed, and it came down to the data analysis component. This led to an investigation into what other fields I could pivot into, which led to my enrolment in my postgraduate diploma. My advice to others is to look at your motivations for wanting to get into a data science role, is it because of the Harvard Business Review tagline of “Sexiest job of the 21st century” or is it because you enjoy working with data to add value to businesses?
What sparked your interest in working with data?
Turning the raw system data into something that can add value is the element that really interests me. Across my career I have had the opportunity to work with raw photogrammetry, lidar, and with population data. This has delivered flood mapping, environmental impact assessments and estimates of hydropower potential. These projects added value to the organisations and clients I was working for. My career has changed the use of geographical data to the use of system data that relates to customers, the challenges are different, but I find the benefits for the organisation to be more rewarding.
How did you come to work in your current role?
I was aware that Bankwest was a leading organisation that really invests in the technology and skills to deliver great experiences for their customers. It was also my first foray into looking at more positive customer analytics that is ingrained into the organisation. It has been a great move and the team and management really embrace what we can do
What sort of projects have you been working on?
Prior to the COVID-19 pandemic, I was building predictive models around some of our products. Now, the focus is on delivering insights focused on supporting the financial wellbeing of our customers.
What tools/platforms do you use in your work?
The standard tool we use for getting the data is SQL, and we use Python and tableau for the modelling and visualisations. There is also some Excel and PowerPoint for the communication of outcomes to stakeholders.
What has been a highlight of your data science career so far?
Reflecting back on my data science career so far I think there are a number of things that are highlights- seeing our team grow from three to thirteen with the addition of two cohorts of graduates is one, they bring so much passion and different ways of thinking to the team. I think the other one is seeing the adoption and weight placed in the outputs of our team by Bankwest, they are really embracing the use of data to make well informed decisions about the financial wellbeing of our customers
What challenges have you faced as a data scientist?
One of the greatest challenges is as a result of being a new team in a new space. The concepts of machine learning are easy in principle, the acquisition of the data and the design of a production pipeline to ensure the outputs are robust week-in week-out is a challenge
What are some of the big areas of opportunity/questions you want to tackle in this space?
Some of the big things I’d still like to tackle is the inclusion of non-structured information based on customer sentiment. We have a wealth of structured system of record data that provides a good insight into customer behaviours and trends, but due to the different systems and data flows we can’t incorporate conversational data that would give an amazing insight into what our customers are feeling.
What excites you most about recent developments in Data Science?
One of the most exciting things is the ability to unpack complex models and add explainability and interpretability to them. The use of neural networks is a scary subject due to the black box nature of them, by having the framework to explain them over the top allows non-technical stakeholders and governance forums to have confidence in the predictions and also show that there are no inherent biases that could adversely affect the use of the predictions.
What does the future of data science look like?
The future of data science is starting to look increasingly fragmented with niche areas developing. We’ve already seen the rise of data engineering roles and I’ve even seen adverts for analytics translators to sell the ideas and outputs back to the business. I think this could be a sign that businesses have been sold the hype of Data Science, but they are struggling to get it to work in production
For people considering a career in data science, what is one piece of advice you would give?
I think one of the core things in people considering a career in Data Science is to understand the whole pipeline. Data Science is often glamorised by the models and the technology that is used but there are many steps before that in sourcing and gathering the data as well as ensuring the predictions are utilised by an organisation correctly.