As a leader in Finity’s insurtech practice area, Marcello Negro has always been interested in applying new and emerging technologies to help solve clients’ problems.
HOOKED ON ANALYTICS
In his early professional career, Negro had the opportunity to work in the application of analytics and cutting-edge machine learning techniques.
This evolved into the use of Finity propriety data and innovative analytics techniques he developed as part of the company’s analytics and pricing team to help start-up clients solve specific business problems.
‘I was hooked, and started actively searching out insurtechs we could help, which led to some of the projects I’m most proud of in my career so far,’ he shares.
A LASTING IMPACT
The few client projects that come to mind include helping start-ups set Day 1 go-to-market prices for new and innovative insurance offerings.
‘Without some of the work that we performed, I like to think that these insurtechs wouldn’t have launched as quickly or as successfully as they did,’ he says.
Having a lasting impact on the insurance landscape still excites Negro, and he plans to continue on this path, working with a wide variety of clients.
LESSON IN RISK MARGINS
Born and educated in the northern suburbs of Melbourne, Negro attended school locally and qualified as an actuary with an honours degree in commerce from Melbourne University as well as a mathematics diploma.
In early high school, he wanted to become a civil or construction engineer but was frustrated by some work experience in the field.
‘I used the calculation for πr2 I learned in maths at high school to estimate the volume of concrete required for a roundabout,’ he says.
‘After plugging 3.1415 into my calculator and getting an answer, I presented this to my supervisor who promptly told me “we use π = 5.0 around here to give ourselves a buffer”.
‘From then on, I vowed to do something a bit more exact, and eventually found my way to actuarial science. I didn’t find out about risk margins until many years later.'
THE RIGHT MIX
Negro says the actuarial profession has just the right mix of interesting maths, interaction with financial markets and human behaviour as well as a healthy business element — all of which fascinate him.
In addition, he says after some early roles in superannuation and stockbroking, insurance really stood out as an industry in which he could do engaging and meaningful work.
‘I’ve been at Finity for almost 10 years and haven’t been disappointed yet,’ he enthuses.
UNPACKING THE BUZZ
In an upcoming ANZIIF webinar, AI and the Future of Insurance, Negro plans to ‘demystify AI and provide a practical definition of what AI is and what it isn’t.’
‘Let’s acknowledge that AI is a bit of a buzz word at the moment,’ he says, ‘but in reality, it’s a concept that has been around since the 1950s and the dawn of computing.’
Negro points to two main reasons that AI has entered popular culture more recently, as well as all aspects of business — including insurance.
‘First there has been an exponential increase in the storage of digital data over the past two decades and second, in the last 15 years, we’ve seen significant advances in machine learning techniques — specifically, deep learning.
‘Whether or not this growth trajectory continues at the same pace is anyone’s guess, but for what it’s worth, I think it will. Needless to say, AI is here to stay.’
The working definition that Negro employs when speaking with clients, colleagues and friends about AI is ‘anything that a machine can do that seems “intelligent”’.
‘This is clearly an evolving definition,’ he says.
‘A chess playing computer was considered AI in the mid-1990s, but now we can all access even smarter apps on our phone that we wouldn’t consider AI.’
To illustrate just how far we have come, Negro also says many insurers are using machine learning techniques that would have been considered at the cutting edge of AI just five years ago.
‘This includes Gradient Boosting Machines (GBM), model ensembles and natural language processing models.’
Just like the chess playing programs of the 1990s, these are becoming more and more part of the status quo.
DATA IS THE DIFFERENCE
However, the key to the advancement of AI as a disruptor like self-driving cars, quantum computing, or the digital economy, is its ability to make sense of the ever-increasing volume of digital data.
‘Modern insurance is based on understanding the risks faced by an insured in as much detail as possible.
‘To do this we look at data collected in the past and try to make sense of it.
‘We not only have more data than ever before, but our data is more diverse, and we need to make sense of it quicker than we ever have. AI includes a family of tools and techniques to help us navigate this.’
But for Negro, AI also represents the infinite possibilities of modern technology.
As an example, he highlights the Artificial Immune System (AIS) as an advanced application of AI based on a relatively new software product developed by the AI team at Finity.
‘The AIS uses a novel artificial network to model the joint probability distribution of a set of data,’ Negro says.
‘This can be applied to a wide range of industries including financial services, insurance, digital advertising and cyber security — to name a few.
‘The network becomes the backbone of an intelligent system able to classify data appropriately to identify issues like insurance fraud, claims leakage and abnormal customer behaviour.’
The product is a specific application of anomaly detection systems, a sub-field of machine learning, which Negro says will soon be used to monitor data that enters an insurer’s system.
The AI acts as a firewall for all areas of the value chain, detecting and responding to incoming data in real time.
‘As a result, risk is reduced and ultimately, the customer experience is improved,’ Negro says.
At the webinar on 16 June, Negro will offer more specific examples of how AI is being deployed across the Australian insurance industry and the impact this has had on businesses.
‘I’ll also summarise a framework for boards and executives to use in assessing AI applications in the insurance context,’ he says.