King Price joins vanguard of insurers using AI for service efficiency
South Africa-based King Price Insurance took the plunge into artificial intelligence and has already seen benefits in fraud detection and claims processing.
By Heath Muchena
King Price Insurance
Artificial intelligence is starting to have a profound impact on the services that enterprises are able to offer, and insurance company executives appear to be embracing machine-intelligence based predictive analytics and automated processes more wholeheartedly than their counterparts elsewhere.
AI could be instrumental in every major insurance decision within the next decade, from detection of fraud to customer service improvements, industry analysts say. Advanced applications are already affecting distribution and underwriting, with policies being priced, purchased, and bound in near real time, according to the Insurance 2030 report from McKinsey & Company.
South Africa-based King Price Insurance is emblematic of those enterprises that have already started to tap AI to generate accurate predictive analytics, optimize business processes, automate routine jobs and allow staff to focus on high-value tasks.
“At the moment, our IT team is focused on improving our business processes using artificial intelligence to perfect our premiums and customer experience, which are the main reasons why clients cancel their cover,” said André Martin, CIO at King Price.
King Price Insurance debuted in 2012 and made a splash in the insurance industry by introducing premiums with costs that declined as cars depreciated, selling its first 100,000 policies within two years.
AI has played a big part in optimizing processes and customer services.
AI to improve claims experience
“Insurers are turning to AI to improve their clients’ claims experience, with a specific focus on claim approvals,” Martin said. King Price has already put AI to use with its Geriatrix system, “developed in-house using Microsoft solutions integrated with data on hybrid cloud, which automates claim approvals and gives our clients a better and hassle-free experience.”
King Price has used the system for a year. “It has been providing a more reliable fraud detection solution compared to manual intervention. This has allowed the business to deliver a faster claims process for clients and enabled staff to focus more on delivering better customer service,” Martin said.
Adopting AI solutions based on solving specific problems in the business helps companies focus on the necessary data sets which helps to achieve set objectives, according to a KPMG in its research study, Powering Insurance with AI. Among other things, it helps enable close collaboration with stakeholders, which is necessary for any business looking to develop use cases to support the achievement of key business performance indicators, KPMG reports.
Building and deploying an AI application requires engineers with data manipulation skills, data domain experts with good data features know-how. Good data scientists with strong understanding of statistics, algorithms and data modelling techniques are key. So are developers and DevOps who are capable of integrating the AI into an application.
AI development requires multiple skills
“To achieve what we have with the solutions we developed in-house, we’ve had to facilitate cohesive collaboration among our developers, technical architects, software architects, and solution architects,” Martin explained.
Martin is not alone in the insurance world. According to research by Genpact, insurance executives are much more likely than their counterparts in other sectors to strongly agree that AI is improving their ability to make more effective strategic business decisions. Redundant processes can create inefficiencies in insurance ecosystems.
“We’ve all been there… spending way too much time on the phone with a call centre agent, answering question after question, listening to an endless list of underwriting requirements. While this model still flies today, it’s fast becoming less popular,” Martin says.
One way AI will make an impact is by optimizing the ratings engines used by insurers.
“Every single insurer has some sort of rating engine which is essentially a function within a billing system that assigns the charging rates to a usage event,” Martin explains. “Rating engines enable design, construction, and testing of algorithms and rate tables in support of an insurance product. These engines are made up of all the data that gives an insurer a competitive advantage in the industry, with a vast majority of the IP coming from years of data analytics and actuarial research.”
According to the International Data Corporation, spending on AI systems will reach US$77.6 billion in 2022 with a significant amount of that investment directed to conversational AI applications such as chatbots and machine learning apps. Such investments in AI technology are expected to save insurers almost US$1.3 billion while also reducing the time to settle claims and improving customer loyalty.
Insurance processes are ripe for disruption
Many companies, however are merely duplicating their existing business model to fit the needs of a more digital platform, typically an app, Martin says. “This usually results in a low uptake within these channels as clients simply end up trading in the time they would’ve spent talking to a consultant, for time on the app,” Martin says.
“While the insurance industry hasn’t found an innovative solution to this yet, AI definitely has the potential to disrupt the traditional underwriting model. In fact, it can be used to shorten this process, while still maintaining the accuracy of risk profiling.”
King Price plans to fine-tune its AI to improve risk profiles, products, premiums and excesses — the initial sums payable by the insured in the event of a loss since they constitute the uninsured portion of the loss.
The company is also in the process of building AI-powered solutions to generate credit scores, Martin elaborated. “Insurers find it challenging to quickly obtain credit scores and, as a result, present clients with their premiums without delay,” Martin said.
“The age-old ‘solution’ is to use a default credit score that either benefits the client or places risk on the insurer. Using AI to get to a ‘smart credit score’ is a more accurate risk assessment solution.
Continue reading for free
Create your free Insider account or sign in to continue reading. Learn more