RFi Performance Analytics Opinion - Australia: Learning from other Industries

I was recently honoured to be able to sit down with a number of incredibly talented individuals at the Menzies Institute. Their goal is to identify the factors that exacerbate and ameliorate the chances of developing diseases by analysing vast amounts of data. While in many ways the subject matter of medical research and finance are completely different I was surprised to find we shared common challenges.

"At its core, we are trying to understand as much about individual customers, portfolios and organisations as possible to gain actionable insights."

The first thing we had in common is that both fields are working with big(ger) data than they have used traditionally. RFi Analytics processes data sets that have literally billions of data points related to the portfolios of Australian and New Zealand financial institutions. At its core, we are trying to understand as much about individual customers, portfolios and organisations as possible to gain actionable insights. Whilst financial institutions have a lot of data on individual customers it is a rounding error compared to the data associated with individuals used in medical research. As well as demographic factors about an individual, they have an individual’s mapped genome. The human genome contains about 3 billion base pairs of DNA that reside in every one of our cells. The resulting data set for a thousand-person study will have over 3 trillion pieces of core data to analyse even before derived data is taken into account. This is big data I was asked what I would do if I was asked to establish a system to analyse that sort of data. I must confess to being a little thrown, but the more we talked it through, the more potential solutions, approaches and technologies, presented themselves. Strategies we use in the world of finance of segmentation, clustering and sampling could all potentially assist; as could cloud platforms. This turned the conversation to data security.

"Data governance, security and protecting the privacy of individuals in line with regulatory and ethical requirements, are front of mind in both fields."

Data governance, security and protecting the privacy of individuals in line with regulatory and ethical requirements, are front of mind in both fields. We spent some time talking about how privacy concerns in both fields had historically hampered the adoption of cloud based platforms. I was able to share the journey RFi Analytics had been on and some of the techniques we use to mitigate and eliminate these risks. Techniques such as the judicious use of use of non-reversible hashes, military strength encryption of data at rest and in transit, two-factor authentication, never storing personal identifiable information and penetration testing to name a few. By simply talking through some of these I think we may have highlighted a potential approach that might allow the Institute to leverage cloud platforms in combination with their internal platforms that would meet or exceed their governance, privacy and security requirements.

Both fields are seeing an increased amount of effort being expended on “data wrangling”. Data wrangling is the process of taking data from one or more sources and transforming and mapping it into a format that is appropriate for a variety of purposes such as analytics. This time-consuming exercise is not only necessary but often requires deep domain knowledge to ensure any resulting data sets and models are valid to help avoid erroneous conclusions. In this instance, our discussion got me thinking differently about how we identify and handle internal data anomalies.

Both fields are coming to terms with new technologies such as machine learning, deep learning and artificial intelligence. We both shared a distrust of “black box” solutions and instead are using these techniques to identify the proverbial nuggets of insight gold that warrant further investigation using traditional statistical techniques. That said, these tools present a game changer as they dramatically speed up the identification of potential insights.

We also spent some time talking through how we show the results of our analysis. The world of finance may not be a world leader in this field but we do tend to apply a lens of persuasion to any visuals - often resulting in a more engaging aesthetic. It was a lot of fun exposing some of the PhD students in the room to some of the more engaging and dynamic visual tools we have at our disposal. I could literally see the analytical cogs turning as they imagined how they would use some of the tools to present their latest findings.

"We both shared a distrust of “black box” solutions and instead are using these techniques to identify the proverbial nuggets of insight gold that warrant further investigation using traditional statistical techniques."

Something else we had in common was a shared belief that no matter how engaging the charts or presentation, the commentary is just as, if not more, important. The associated commentary provides the shading, context and expert opinion that helps the end user of the insight focus on what the analyst has identified rather than being left to draw their own conclusions.

"The associated commentary provides the shading, context and expert opinion that helps the end user of the insight focus on what the analyst has identified rather than being left to draw their own conclusions."

This one short meeting spurred a couple of interesting ideas for all participants. By stepping outside the comfort of my domain knowledge I gave myself the chance to learn something new. The core data and analytics skills across both disciplines are very similar, but the differences in the data and objectives led to a surprising, challenging and genuinely interesting experience. I would like to take a few words to say thank you to the Menzies Institute for opening my eyes to the benefits of cross industry discussions around data and analytics and the potential benefit of hiring resources with the requisite technical skills from other industries. I hope they have every success in their pursuit to deliver internationally significant medical research leading to healthier, longer and better lives.

 

Upcoming Events
26
Oct
21
Australian Banking Innovation Summit 2021
Sydney, NSW, Australia
See all upcoming events
map4
Subscribe to receive insights delivered straight to your inbox
Latest news, unbiased expert analysis and insights across banking and finance