Artificial Intelligence

Eric Chalmers tells us how brokers can use AI to automate some of their workflows. (Runtime: 3 min, 34 sec)

Size: 5.2 MB,   Time: 3 min, 34 sec   Text transcript  Download

broker on the go podcast


Eric Chalmers, director of data science with Surex Direct, explains how brokerages can leverage artificial intelligence and machine learning to take their business to the next level.

Related Article: Surex Direct’s new hire takes on AI, machine learning


Subscribe to our Newsletter 

Text Transcript

Eric Chalmers, director of data science at SurexDirect.com.

Artificial intelligence and machine learning can benefit brokers in a number of ways. For example, they can make the brokerage more efficient by automating repetitive tasks or decisions like reading and categorizing documents, doing data entry and data lookup—the kinds of routine tasks that we accept as part of life, but actually eat up a good portion of our day.

Using AI to automate these kinds of tasks improves operational efficiency and makes staff happier too. With AI systems automating lower-level tasks, staff are free to focus on higher-value tasks that grow the brokerage or improve the level of service it provides.

Another thing artificial intelligence and machine learning are good at is finding patterns in data. As a broker, if you’ve been collecting and storing data on all the quotes you generate and all the policies you manage, machine learning algorithms can analyze that data and pull out insights that could be valuable to you or to other players in the industry.

And brokers can collect different kinds of data than insurance carriers can, which means they may be able to answer some questions that carriers can’t answer. For that reason, I think that AI could actually start to change the role that brokers play in the industry.

Something else to consider is that artificial intelligence is more accessible now than it’s ever been. If it doesn’t make sense to hire your own data scientist or AI expert, you can find data science consultancies who will work with you, though the costs will be high.

If you don’t want to do that, you can probably find specific AI systems for sale or license. At Surex, we’re open to licensing technologies that we develop, because we feel that building up the broker channel is good for the industry.

There is a sentiment out there that AI is going to put people out of work. It’s related to this Hollywood characterization of AI as something pernicious, but those perceptions are just erroneous.

Intelligent software is a tool that can extend our abilities, enhance our decision-making and let us offer more accurate and higher quality service, and Canadian brokers should make adoption part of their business strategy.

Surex Direct has deployed a few AI-based systems already and we have several more in various stages of development or planning.

One thing we’ve done is use AI to automatically read and sort documents. As a broker, you handle a lot of paperwork, and we’ve found that reading, sorting and processing all those documents eats up a lot of time for our support staff.

This new AI system we’ve developed sorts documents automatically. What’s interesting about this system is that it learned how to sort documents by watching staff members sort documents. I’m sure every brokerage has their own unique workflows, but the intelligent system is able to infer our procedure by simply observing our staff, and it will also adapt intelligently when things change.

In the case of billing-related documents, we have another system that can read those documents and extract the important information. It automatically does the data entry and data lookup related to that billing issue, and in most cases, it also notifies the customer of the issue before handing things over to our billing team.

Automation is saving each of our billing staff up to 90 minutes per day. The system handles more than 90% of documents successfully, and the cases that it can’t handle are referred to staff. It learns from them how to handle similar cases next time.

So those are some examples of how we’re using AI for automation. We’re also using it for decision support and quality assurance.

For example, we’re building an intelligent auditing system that will learn what the appropriate coverages are for a given situation. So if a broker forgets to add, say, a limited waiver of depreciation endorsement for a new vehicle, the system will recognize that as unusual and refer that case to staff for review.

Copyright © 2018 Transcontinental Media G.P.
Transcontinental Media G.P.