AI and the future of insurance and underwriting
UTS Advanced Analytics Institute (AAi) is collaborating with ANZ-OnePath, one of Australia’s leading providers of wealth, insurance and advice solutions, to harness advances in machine learning, data science and predictive analytics to do things differently.
This article originally appeared on the CeBIT Australia blog.
Machine learning algorithms that give computers the ability to analyse business processes and big data are discovering new opportunities and insights for improved processes and better service, and UTS and ANZ-OnePath together are exploring opportunities for artificial intelligence (AI) an algorithm with human decision-making capability and data science the extraction of knowledge and insight from data to transform the insurance application process.
They are combining their professional expertise in analytics, and in insurance underwriting, to examine ways to significantly reduce processing times, improve quality and offer cover to more Australians. The AAI analytics experts will examine the huge amounts of data that is collected from customer questionnaires to identify ways to reduce a time-consuming manual process which can leave customers waiting weeks for approvals, or even walking away, uninsured, due to frustrating delays.
The traditional underwriting process asks customers up to a 100 possible questions; these are standardised and not personalised or tailored to their particular situations. The underwriter then needs to collate them manually to determine an application – or even seek further specific information, such as medical reports, adding extra stages to the process.
ANZ-OnePath wants to simplify this process, make it easier and more accessible, and improve the customer experience over the whole life cycle of an application for cover.
And artificial intelligence and machine learning can be applied to do just that.
WATCH: UTS and ANZ OnePath – AI enhanced underwriting
[text on screen]: Improving insurance underwriting using artificial intelligence
[text on screen]: Artificial Intelligence – A computer algorithm with human decision making capability
[text on screen]: OnePath had a challenge…
[text on screen]: Up to 100 key questions involved in assessing a customer's application
[text on screen]: Up to one month wait for the customer to receive a response
[text on screen]: Questions were designed with professional input but not based on data insights
[text on screen]: The UTS Advanced Analytics Institute devised a solution for faster insurance processes
[text on screen]: Data science explored years of business data to identify the most useful questions
[text on screen]: Based on those insights, machine learning reduced key questions
[text on screen]: Machine Learning – Algorithms that give computers the ability to analyse business processes
[text on screen]: Result? Insurance outcomes in seconds
[text on screen]: Better on line customer experience. More streamlined business process. Offer cover to more Australians. Enables ANZ OnePath to be a world leader in AI drives underwriting.
This new partnership has come about in recognition of AAI’s capability in big data, data sciences and data analytics, and its experience in the financial technology sector. This collaboration is the latest by UTS AAi in developing innovative analytics solutions for the financial technology sector.
We aim to be the leading research group in applying data analytics and AI in FinTech across various wealth sectors such as insurance, superannuation and investment portfolios
Associate Professor Guandong Xu
“We have developed a four-step methodology for working closely with our partners to build domain knowledge and help them to be more agile and develop quick outcomes.”
This 4D methodology – Discovery, Design, Delivery, Deployment – is used by a team of twenty to understand the business, the challenges it faces, and, on a case by case basis, map specific problems to develop a data-driven concept design.
ANZ-OnePath chief underwriter Peter Tilocca said using a data-driven solution is a clear opportunity to create a more personalised, efficient service with improved quality assurance for customers applying for insurance.
Key questions we began with:
- Can the questions we ask of applicants be tailored, personalised and reduced?
- Could we use AI and data science to develop an automatic risk engine to quickly and consistently identify if an insurance application is standard, or has an exclusion or loading applied?
- Could we use AI and data science to improve quality assurance processes internally for ANZ-One Path?
“The answer to all is yes and we now see how AI can develop a more reliable application and assessment process and also provide our advisors with a differentiated service that supports them in building trusted relationships with their clients,” he adds.
With ANZ-OnePath, Associate Professor Xu and his team used data science to explore the historical data across the insurance lifecycle process and combined this with AI to develop an Automated Underwriting Risk Engine.
“An intelligent underwriting model will harness machine learning to provide the opportunity for insurers to develop more efficient and reliable assessment processes. We will involve client behaviour modelling, text mining and natural language programming, along with social and predictive analytics that can add value in the insurance sector,” he says.
“We discovered that many questions do not provide useful insight to determine if an application should be accepted, have an exclusion, loading or be declined. Data science applied to intelligent underwriting shows that just use 10% of questions asked can achieve the same underwriting outcomes as using all the questions.”
The virtual underwriter will determine if a loading or exclusion needs to be applied and will only process applications that have been accepted by the customer for underwriter confirmation. It will save time, and saves ANZ-One Path time and increase operational efficiencies and product quality.
The automated system’s quality assurance structure can also evaluate all decisions (Virtual and Human) and generate quality assurance reports for managers to review. The unique advantage of AI is in its adaptive and self-learning capability, absorbing new data and improving over time using aggregated data and external knowledge.
For ANZ-OnePath this is the future of insurance and underwriting. “We want to be a world leader in AI underwriting, and Australia’s preferred wealth and life insurance company, delivering a more tailored, customised and improved on-line digital experience for a wider base throughout Australia,” says Mr Tilocca.