Leaders in Tech: NewDay Senior Specialist in Data Science, Donal Simmie
At NewDay, it is our priority to help customers be better with credit. Our Data Science team plays a big role in this, as Donal Simmie explains.
“We have a great opportunity to have a positive impact on our customers. We’re focused on responsible lending.”
In a previous Data Science role elsewhere, Donal spent far too much of his time stitching together databases and cleaning up data. “It was probably 80:20,” he recalls. “Only 20% of my time was spent working on valuable analysis — testing hypotheses and uncovering insights.” When he joined NewDay he was determined to reverse that ratio, so that the team could focus much more on adding value.
At that time, the new Data Science team were in a rare position. They started with a blank canvas and they had the chance to build a data platform their way. By making some smart design decisions, they’ve freed themselves from a lot of those low-value activities.
NewDay, with around 1000 employees, is relatively small to have the luxury of such a powerful data platform. At the same time, there are big advantages to being the size we are — it’s a sweet spot. Data Scientists can see how the whole business operates. They’re not cogs in a much larger machine; they get the opportunity to see the results and benefits of their work.
Donal himself brings a background of social network analysis. This is unusual in Fintech, an industry which tends to aggregate data and overlook individual behaviour. Instead, by treating transactional data like a network, Donal and his team are able to put together new behavioural segmentations. These are much more powerful than traditional demographic approaches, and much more accurate than qualitative insights from small scale surveys.
Working fast, working smart
The team also get to work on an incredible variety of projects. Some of the recent work includes: simulations of debt over time; finding the best time to call customers; improving the response to various marketing approaches; and selecting the best affordability criteria. So employees have to be generalists with the ability to specialise in a new field very quickly. They can get a lot of experience in a short space of time.
Of course, our Data Scientists are very talented. They prefer to work fast — to extract data and test hypotheses within minutes or seconds rather than hours — so that they don’t lose focus by constantly task-switching. At the same time, their work needs to be rigorous and well thought through. This means that they need to be familiar with Python libraries, for example. Indeed, they’re innovating strongly in the Python data space and plan to make some projects publicly available in 2019. They’re also building a new Retail Analytics product — they can’t say much more about that yet, but they can say that it uses cutting-edge technologies and innovative Data Science techniques. On the software development side, they’re big users of the Apache Arrow project which enables fast data transfer between all of the systems they use in the Big Data ecosystem. In terms of modelling they’re starting to leverage probabilistic programming more and more; projects like PyMC3 have made this adoption much easier, as most of their work is implemented in Python.
At NewDay we see great potential in new thinking. We’re proud of our linkup with Imperial College — one of the top computing colleges in the world. We were one of seven companies featured in their 2018 Data Challenge (link: https://medium.com/@newday/newday-ic-data-challenge-2018-4032feb11ea9), and we sponsored their AI Hack in November (link: http://aihack.org/).
Another sort of network
Our Data Scientists are also naturally entrepreneurial; they spend time networking. Donal cites the example of a recent new starter. “Without any prompting, he found a list of the key people around the business. In the first week, he’d set up meetings with all of them, and he’d identified an opportunity to improve direct mail response.” There are plenty of opportunities around the business to make even better use of data. Different teams in NewDay use different segmentations; for example, there are separate dedicated segmentations for risk, for marketing, and for customer journeys. They also use different models to help them achieve their various objectives. When the Data Scientists spot an opportunity, they need to translate it into something they can achieve, and then deliver.
At NewDay, we are creating a positive culture and environment to help get things done. We believe in co-located teams which collaborate and interact regularly. The mood is open and trusting, and we’re not hierarchical. We hold agile sprints and retrospectives. When something isn’t working we take the time to fix it properly. Individuals have autonomy but are held together by strong social bonds.
Does this sound like a workplace you’d thrive in? If you’re a Data Scientist with an outstanding track record, and you’d like to spend less time cleaning data and more time adding insight, we’d love to hear from you. Find out more about working at NewDay or browse current vacancies now.
