BAE Systems

The burgeoning need for data science has created a banking skills gap. In his latest blog, David Nicholson discusses the use of ‘citizen data science’ as a means to mitigate this gap by empowering a wider pool of employees to engage with data science workflows

Tuesday 8 June
Read time: 3 mins
Applying citizen data science
Banks are increasingly turning to data science in business application contexts involving complex data sets that contain events of interest, which are typically weak and uncommon. The detection of fraudulent and financial crime patterns are primary examples. In this setting, data science technology, drawing on algorithms and methodologies from the fields of statistics, optimisation and machine learning can be a powerful enabler for enhanced discovery and detection.
In the banking sector, driven largely by the growth and complexity of data, being able to unlock accurate business intelligence using data science is central to a number of applications. While these applications include fraud and financial crime detection, they also extend to a range of other front office and back office applications within banks.
This burgeoning need for data science risks creating a skills gap within banking, not just in terms of recruiting qualified data scientists, but also retaining them on projects when there are multiple demands from other projects. In this blog, David Nicholson discusses the use of ‘citizen data science’ as a means to mitigate this skills gap by empowering a larger resource pool of data professionals to engage with data science workflows through appropriate technology and tooling. 

Role of citizen data scientists

‘Citizen data scientists’ was a phrase coined by Gartner to describe individuals within an organisation who could build data analytics and modelling workflows “but whose primary job function is outside the field of statistics and analytics”. Such roles require intuitive technology and tooling to create these workflows so that users can engage with it without the need to be data science experts.
The use of intuitive data technology and tooling would provide an opportunity for banks to plug the data science skills gap by drawing from its more abundant data savvy workforce. It would enable them to develop and apply advanced data science workflows, whilst remaining within their comfort zone, using tools and technology that have a familiar look and feel that condenses the deep data science details.
While citizen data scientists are not a direct replacement for skilled data scientists, their utilisation on projects can reduce the resource pressure on data scientists’ time. It will still be important for such roles to check-in with a data science expert during a project, to seek guidance around key workflow decision points or to assist with the interpretation of modelling outputs, for example.  
Applying citizen data science blog

Core challenges to citizen data science

In practical terms, there are several ‘pain points’ that must be overcome to empower a citizen data scientist to work rapidly, comfortably and confidently through the creation of an advanced data science workflow. Data science skills and experience mainly come into play when machine learning (ML) models are trained, optimised and validated. A citizen data scientist should not have to wrestle with Gradient Boosted Tree code and setting model parameters (for example) – the tooling should simply self-serve a menu of models to the user and automatically optimise parameters if required.
Data science tooling needs to package solutions for use by citizen data scientists in a way that is neither so simple to use that it compromises on insights or accuracy, nor too complex to use that intended users find it too difficult to use and need expert data science help. The challenge is to hit the sweet spot that places technology and tooling into the hands of citizen data scientists that is capable of generating accurate and useful results, but that also minimises friction. This is key to resolve this challenge and get the balance right.
Banks need to demonstrate compliance both to internal assurance teams and to regulators. The risk is that ‘dumbing down’ on data science to mitigate a data scientist skills shortage can result in a weakening of banks defences against fraud and financial crime. That is not an option, but conversely the packaging of data science must reflect a broader user base than just data science experts.

Data ethics cross roads

Building on the previous point, data science solutions employed by banks must stand up to scrutiny in terms of fairness, accountability, transparency, and ethics (FATE) requirements imposed by policy and regulation frameworks within the banking sector. By abstracting data science details inside a shiny tool or user interface, some of the nuances of data science models and their configurations that could potentially amplify bias and promote unfair decisions may be overlooked.
It is important therefore that tooling aimed at citizen data scientists in banks, even though on the surface does not expose users to deep data science details, has thoroughly documented those details and is open to interrogation at a deeper level to satisfy the FATE requirements.
Given the data scientist skills gaps and the rise of data-driven applications, banks will need to evaluate resource, technology and tooling needs to manage this situation. Empowering citizen data science should be a key aspect of that consideration, building on global progress in the democratisation of data science. Balancing the interplay between citizen data science and deep data science, between human and machine intelligence, and between performance and FATE will all influence what the best outcome will look like and how it will work in practice. 

Other recommended content

Do you want to access more blog content from our Financial Services experts?

More blog content
banking sign up

Sign up to get the latest industry intelligence and insight

Stay on top of the latest news, forthcoming webinars, new podcast episodes and upcoming trends in Banking, Insurance, Data and Cyber by signing up to our BAE Systems Insights series. Hear from industry experts sharing their views on hot topics and new technologies. Delivered fresh to your inbox. 

Get in touch with our experts today


Americas  +1 720 696 9830     |     Europe, Middle East  +44 (0) 330 158 3627     |     AsiaPac  +61 290 539 330