TEXT NEUS MARTÍNEZ| PHOTOS MAPFRE, ISTOCK
The potential of artificial intelligence (AI) goes far beyond the audiovisual content that Netflix can recommend us on the basis of what it has learned about our tastes, or the skills with which virtual assistants like Alexa can make life easier for us in a connected home.
In the insurance sector, the use of artificial intelligence is blurring the boundaries of the industry and accelerating innovation. This group of technologies is set to lead a transformation which entails: enhancing the customer experience with current products and services; developing new products and services that add value to the increasingly personalized demands of policyholders; improving operational efficiency in business processes; preserving the integrity and security of data and operations; and taking service quality to the maximum level.
All of this is taking place in a world that was already digital and has become even more so as a result of the COVID-19 pandemic, within an environment in which omnichannel distribution and immediacy of communications and transactions are now the norm. Customers have revamped their expectations over the past year, as a result of a reassessment of their needs and a prioritization of values. The whole playing field has changed and, therefore, the organizations and the role they play have likewise changed. The insurance industry must be ready to provide a personalized service to customers who are always connected, increasingly demanding and, more than ever, aware of their needs. It must satisfy requests precisely and immediately and, moreover, create a closer digital relationship. This has been the case throughout the recent health crisis where it has acted as “a social shock absorber, serving its clients in these extremely difficult times and doing so efficiently with well-trained, accessible personnel,” declares José Antonio Arias, Group Chief Innovation Officer at MAPFRE.
AI: relieving people, benefiting people
Artificial intelligence is present in almost every area of our lives and MAPFRE is aware that this technology forms part of the solution to many of the challenges currently facing the insurance world. That is why the company set itself the goal of incorporating it, so as to swiftly transform the experience of current and potential customers, and create next-generation products and services. The goal is to offload some of the complexity – to date handled by individuals – onto this emerging technology.
Driving innovation and benefiting from the power of sophisticated analytical techniques – and technologies such as the one specifically addressed here – proves possible thanks to the efforts of a team of professionals working in different areas of MAPFRE: operations, business development, technology and innovation, among others. In short, these are people who know the business, processes and AI applications that can help us reinvent and enhance our services and products, such as ‘deep learning’ (deep neuronal networks) or ‘machine learning’. Together, these teams and technology are making pioneering services and products like the following a reality.
Claims for damages resolved in minutes
Filing an insurance claim is often a complicated process for customers, who sometimes feel overwhelmed by the amount of information required, as well as the uncertainty re-garding how long the whole process will take or how it will all turn out. Providing an im-mediate, accessible, transparent experience was the objective for MAPFRE; this was achieved thanks to an intelligent automated solution, available at any time, that reduces from days to a question of minutes the resolution of a claim for water damage in the ho-me in a precise, reliable fashion.
"THE ARTIFICIAL INTELLIGENCE SYSTEM IS ABLE TO UNDERSTAND WHAT OUR CUSTOMER NEEDS"
The magic behind the result: machine learning
According to Esther González, an Innovation expert: “The artificial intelligence system is able to understand what the customer needs. To do this, first of all it interprets the user’s natural language and then it analyzes, processes and extracts documentation from the claim file, whether it be an invoice, a loss adjuster’s report, or other documents necessary for processing the claim. This information and other variables are taken into account by the cognitive engine when determining the outcome that corresponds in each case.” It seems magical, but it is called machine learning; this is the automation technology that opens the door to replicating this project in multiple use cases, not just in claims processing where it has already been applied, but also in other business processes.
This solution employs a chatbot as the user’s interlocutor. It is easy to instruct customers to use this method. This is how Daniel Almodóvar, Innovation Delivery Manager, explains it: “If the claim comes through the call center, you can send the customer an SMS to their cell phone for them to access the chatbot in real time; and, if the request arrives by email, you can include a link in the reply. The idea is that, whatever the channel, you can connect the user to the conversational assistant quickly and easily, taking into account, moreover, the fact that the customer has the possibility of switching to human interaction at any time. This automation process is not intended to be obligatory, but rather beneficial.”
The user experience is another element that is afforded the maximum care: “While they are complex technologies, as regards both information collection and the documents and policies, we always gather data in a simple manner and communicate with clients in a transparent, jargon-free fashion,” Esther González remarks.
Process automation applies to different business areas and covers all the stages that customers and users pass through, from managing already active policies to the first step of insuring a vehicle, which includes the appraisal of its condition, to cite another example.
IA en MAPFRE
DAMAGE APPRAISAL USING AI SO TAKING OUT INSURANCE IS JUST A CLICK AWAY
Can an algorithm distinguish dirt or a stain from damage in a photograph of a vehicle? The answer is yes, thanks to deep learning, a technology employing algorithms drawing on neural networks that attempt to mimic the behavior of the human brain.
THE USER TAKES CONTROL OF THE PROCESS
In this new approach to the experience of taking out automobile insurance, the customer takes pictures of their vehicle, while the damage recognition tool processes the images and verifies its condition so as to prepare the policy. Previously, the insured had to request in advance a damage appraisal of the vehicle or, failing that, go in person to the appraisal center, confirm any damages and take out the policy. With this new solution, the customer takes control and can handle everything from their cell phone. Within minutes, the damage appraisal is complete and the policy can be taken out with the desired cover options.
The project’s artificial intelligence system has achieved a very high level of precision when analyzing the pictures that customers send in of their vehicles. According to Mireia Rojo, an Advanced Analytics expert and Pedro Sacristán, Artificial Intelligence Lead: “There are multiple factors involved in achieving that outcome, but the experience of the team with complex algorithms stands out. The quality data used for algorithmic training – thousands of photographs of vehicles that have been processed and labeled – also proved really important. And, without a doubt, persistence in the training iterations which enabled the results to be refined.”
As in all innovation projects undertaken at MAPFRE, the objective is to facilitate life for our customers, providing new tools so as to gain in efficiency, precision and speed throughout every process, all in benefit of the user experience. This project is one more response in the move to align the insurance world with a new digital environment that calls for immediacy in transactions.
Likewise, artificial intelligence is a potential generator of projects that are rather more than a mere redesigning of a process or procedure. It is likely to provide innovation with a social impact.
PERSONALIZED EXPERIENCE IN THE FUTURE
As regards safe driving, thanks to the use of artificial intelligence combined with geolocation, MAPFRE has designed a solution that radically changes the experience for drivers by creating positive stimuli and educating them in safer habits.
This next-generation service — complementary to conventional navigation systems — cautions drivers about potential road hazards ahead by means of non-intrusive warnings and alerts, while suggesting safer alternative routes. In the future, these notifications will be tailored to what the driver considers relevant, and will incorporate additional information about the road conditions and warnings about the policyholder’s driving, thus offering a fully personalized experience. An example of how innovation and the use of smart technologies have a social impact and can con-tribute to enhanced road safety in our cities.
This project uses an artificial intelligence model, deep learning, which analyzes multiple types of data drawn from different sources, such as road accidents, traffic and weather conditions, and the particular characteristics of the route (curves, inclination or condition, among other varia-bles).
SAFE DRIVING: WARNING OF ROUTES WITH A RISK OF ACCIDENTS
Thousands of traffic accidents occur every day around the world, with unique circumstances in each case, but all of them have something in common: external factors play a decisive role. A poorly lit road, high traffic density, sun glare, or a curve that is tighter than expected are all possible hazards.
Accompanying drivers on their daily routes and warning them of dangers ahead, predicting and informing of these risks dramatically transforms their driving experience.