Allstate NI Data Scientist Deborah Hunter outlines her role, with tips for those considering a career in data science.
Deborah Hunter is a Data Scientist in Marketing Analytics (MARA) at Allstate NI in Belfast. She has over 10 years of research and lecturing experience in Physics and Applied Mathematics. With a PhD in Astrophysics, Deborah has worked on many data analytics and modelling projects in this field as well as in Medical Physics at Queens University Belfast (QUB).
More recently, Deborah has applied her academic knowledge and teaching skills to the fields of big data analytics and predictive modelling in Consumer Analytics.
She is currently delivering Allstate's Community of Practice (CoP) training courses in Data Analytics to Allstate employees in NI and beyond.
Allstate NI has provided Deborah with opportunities to apply her practical modelling skills and ML techniques to meet the evolving challenges facing data scientists in the Insurance Industry.
Sync NI’s Niamh Campbell sat down with Deborah to discuss data science at Allstate, and how people from all disciplines can get involved.
There are many data scientists with a physics or statistical background from QUB in our Belfast office. The research skills, coding and statistical side of astrophysics makes it very easy to transition into data science.
My PhD involved statistical data analysis as well as the manipulation of optical and near-infrared telescopic images. I used my IT and mathematical skills to model the spectra of a nearby supernova and to determine the chemical composition of the progenitor star.
One aspect of astrophysics which I loved was collaborating with international teams of research scientists. I wrote proposals for observing time on the Gemini North Telescope in Hawaii. When working with telescopes at the Roque de los Muchachos Observatory located on the island of La Palma in the Canary Islands, I just missed Brian May from Queen when I was out there! I attended conferences in Italy and learned about data analytics processes at the European Southern Observatory (ESO) in Munich, Germany.
My role in Allstate has seen me collaborating with Data Scientists in the U.S and India. I was flown to our Head Quarters in Chicago after only working in our Belfast office for six weeks!
Although I enjoyed teaching and inspiring talented young people, I did miss my research days at Queen’s University. It was when I went on an international teaching programme with CERN in 2015 that I realised I wanted to make the career transition.
As I had taught for a few years I knew I had to go back and learn coding and machine learning (ML) in R and Python as I didn’t learn these in my undergrad. Data Science is such a fast moving and complex field that it is important to be proactively upskilling and extending your experience.
Then when I felt comfortable I applied for data science roles. Physics graduates have statistics, coding and modelling experiences. Having those basic skills really helps.
When I first started working in Allstate I informed them of my teaching background. My work goes between developing predictive models for marketing purposes, such as determining effectiveness of digital marketing campaigns as well as customer segmentation. Our propensity models predict every household with a probability score which can be used for targeted marketing and personalising our customers’ preferences and needs. I can work on my own initiatives and explore new algorithms. I also teach data analytics courses in the Belfast office to Allstate staff. American colleagues can watch the courses internally through Allstate TV.
I also supervise Queen’s placement students studying towards their MSc in Data Analytics. Students typically spend 3 months working on a Data Science related project within Marketing Analytics.
To date, my colleagues and I have delivered to approximately 400 employees throughout Allstate NI (Belfast, L’Derry & Strabane) and have hosted several successful events since first forming in late 2017.
I find Data Science to be a rewarding and fulfilling career as I now have the perfect balance of teaching, collaboration, model development and research.
Allstate encourages progression and taking the initiative to develop ourselves professionally. We have regular group sharing sessions in Belfast because we like to keep up to date with the latest ML techniques and technological developments. The Chicago office also has ‘lunch and learn’ and ‘decoded’ sessions twice a week, whereby over 150 people join Skype sessions to learn about best practice and share experiences in data modelling. The company is great in ensuring we are learning and bettering ourselves all the time.
We have discovered that there is a new documentary called “Data Science Pioneers – Conquering the Next Frontier” which is currently being premiered in London. We’ve planned to hold a screening ‘popcorn movie evening’ in the Belfast office to watch it. We also often have meet-ups open to the public called ‘Everything Data’ which relates back to teaching courses. We regularly host these events and invite guest speakers both externally and internally to share their knowledge, which will hopefully inform more people outside of Allstate about data science and what our team here in Allstate does.
People assume data scientists are stereotypically just good at maths, logistics and coding away in a corner, but data scientists can be quite creative people. A Data Scientist requires a good balance; a creative combination. I myself have a passion for photography and art. We don’t all do the data processing ourselves, as most of our Marketing Analytics team are data engineers and do that too.
Data scientists are curious people who enjoy uncovering new patterns in data, and developing innovative products, services and business models. The novelty in the processes that frame our work, the unpredictability in outcomes and the space for interpretation in the project all require creativity.
Story-telling is also important at every stage from data exploration and feature visualizing to model creation. Some favourite story board tools used at Allstate NI include PowerBI by Microsoft and Tableau. When narrative is coupled with data, it helps to explain what’s happening in the data and why a particular insight is important. When visuals are applied to data, they can enlighten to insights that they wouldn’t see without charts or graphs.
Allstate’s Marketing Analytics (MARA) group delivers analytic-driven insights to solve complex business challenges across the company. The growing team of Data Scientists and Data Engineers together build products that create insights to help business leaders, agents and customers make smarter and more informed decisions. The Data Scientists use mathematical, statistical and machine-learning techniques in the research and development of predictive models and decision-making tools.
This interview first appeared in the Autumn 2019 edition of the Sync NI magazine.
