Some brokerage staff members could be spending hours each day just searching for information in their carriers’ underwriting manuals – time that could be put to better use, the head of a technology vendor suggests.
Some of the longest underwriting manuals are 1,500 pages in length, while most have well over 100 pages, said Joseph D’Souza, founder and CEO of Kitchener, Ont.-based ProNavigator.
Underwriting manuals include information on a carrier’s coverages, risk appetites, limits, deductibles and rules around the types of policies they write. Often each manual has its own tables, and each carrier has its own way of setting them up, said D’Souza.
“As a brokerage you could be dealing with 10 different markets and you could get 15, maybe 20 updates in a month. And it becomes extremely difficult for brokerages to keep up with the changes that are happening.”
ProNavigator’s technology uses natural language understanding that is intended to absorb information from lengthy PDF documents (such as underwriting manuals) and let workers ask questions, D’Souza told Canadian Underwriter recently. He was speaking in the context of a recent announcement about a $5.6-million round of funding.
The round of financing announced Sept. 2 was co-led by Luge Capital, a fintech venture capital fund with offices in Montreal and Toronto. Luge Capital’s funding providers include Industrial Alliance, La Capitale, Sun Life and Desjardins, among others. Luge Capital’s co-investors of ProNavigator include GreenSky Capital MaRS IAF, iNovia Capital, BDC and CIBC Innovation Banking, ProNavigator and Luge Capital said Sept. 2.
ProNavigator’s new funding will be used in part to expand in the United States, where it already has a Florida office. Its target markets include brokers, underwriter,s and claims professionals.
ProNavigator trains its artificial intelligence models to understand the language of insurance, said D’Souza.
That artificial intelligence is built using insurance manuals, as well as thousands of queries workers have made about P&C insurance. For example, the AI would know that if a broker asks a question about SBU, that could mean sewer backup, while GRC would mean guaranteed replacement cost.
Brokers may have to log into different carriers’ portals to find information because it is frequently updated and changed, D’Souza said.
“In a world where we expect things to be done in seconds, and definitely customers expect things to be done in seconds, we saw an opportunity to save productivity time by using a tool that can help retrieve information quickly and understand search queries and questions that are being asked,” D’Souza said.