Insurance
ANZ launches underwriting engine to market

ANZ's life insurance arm has released its predictive underwriting technology following a successful soft launch.

The big data-powered underwriting engine will allow customers applying for OneCare insurance to complete personal statements 30% faster.

ANZ chief of underwriting Peter Tilocca said the number of medical questions has been cut from 38 to 11. What used to take 25 minutes has been reduced to 15 minutes from start to finish.

Entering this project, Tilocca said the aim was not to slash the number of questions but to ask the right questions.

Three years in the making, the collaboration with University of Technology Sydney's (UTS) Advanced Analytics Institute has found meaningful correlation between medical history and demographics, and made the application process simpler.

ANZ completed its proof of concept stage with tradies in mid-April, analysing the data of plumbers, carpenters, electricians and builders aged between 20 and 40-years-old.

ANZ head of insurance operations Jay Tutt said the underwriting process is somewhat "dreaded" as soon an applicant selects 'other' in the questionnaire.

This is automatically flagged with the underwriter who would then work behind the scenes with the financial or risk adviser and client - this requires a lot of manual work, he said.

Tutt said the end game is to minimise the laborious work and give clients a policy in real time.

The technology is expected to reduce underwriting referral triggers by 2300 per year. About 1000 medical conditions have been simplified and made search-friendly in the engine.

UTS Advanced Analytics Institute professor Guandong Xu said: "Using AI to simulate people's decision making process, we have shown how AI has used big data, natural language processing and predictive modelling, to revolutionise a huge process such as underwriting."

In September, the underwriting system won the Best Industry Application of Data Analytics and the Best Industry Application of Artificial Intelligence at the Data and AI Innovation Awards held in Sydney.

Read more: ANZGuandong XuJay TuttOneCarePeter TiloccaUniversity of Technology Sydney
Link to something iVvWLRWy