Company increased the number of frauds identified by 43%
Since 2016, HDI Seguros has restructured itself to optimize fraud prevention, with the aim of better identifying fraudsters and reducing cases of false positives, which affected the customer experience. To do this, it used fraud prevention solutions from SAS, a leader in AI and Analytics, and committed to a journey of continuous evolution. As one of the results, from 2021 to 2022, HDI Seguros increased the number of frauds identified by 43%.
Leandro Bordon, director of operations and claims at HDI Seguros, explains that the changes made in the area in recent years were extremely important for the company to have the people, processes and technology employed to have faster, more efficient and flexible processes, to face the constant evolution of fraudsters. “If we don’t know how to identify a fraud action through accurate assessments, we won’t be able to reduce bottlenecks and waste in the company”, he points out.
However, there were many challenges at the beginning of the project, including defining what the analytical decision engine would be, in accordance with internal rules and the creation of predictive models combined with business strategies, starting in the automobile vertical. Additionally, there was a need to scan the entire database to correct existing anomalies.
“SAS brought a technological tool that helped us analyze external and internal data for better accuracy”, says Bordon. “We are able to look at the relationship network of fraudsters and customers with more intelligence, as it involves interconnections of several elements in a complex way, such as location, contacts, credit score, damages reported in the claim and other information.”
Bordon highlights that SAS' anti-fraud solutions also helped HDI Seguros understand key factors in the prevention process, such as risk propensity, cases that will bear fruit (i.e., types of fraud they should focus on), in addition to enabling real-time decision making.
Reducing false positives was another element of attention to improving the customer experience. And, in addition to reducing cases, the insurer reduced the investigation SLA and evaluation time, factors that also contribute to optimizing the insured's experience.
In 2019, with the good results in the automobile vertical, the company expanded the fraud prevention project to the residential area (property), consolidating a more agile fraud identification process in the company.
“Those who act in bad faith reinvent themselves every day, so we need to be prepared to identify and stop different types of fraud quickly”, adds Bordon. “In this sense, we will continue on this journey of evolution to always have the best possible operational intelligence.”
International reference
The successful project became an example for foreign markets and, therefore, was presented by Leandro Bordon at the SAS Innovate event, which took place in Orlando, in the United States, in May. SAS's main global conference of the year brought together experts from around the world to present trends and success stories in the application of AI and Analytics in various sectors.
Ricardo Saponara, leader of the fraud and financial crime prevention practice for the Americas region of SAS, highlights that among the factors that made HDI Seguros successful in using SAS technology was the understanding and connection of technology with business strategies , investments focused on results and training to enable organic growth, thus reducing the total cost of ownership and maintenance.
“The entire team was trained to extract the full potential of technological solutions, thus being able to create, monitor and adjust fraud prevention strategies independently, maximizing the accuracy of the program on a constant basis for HDI”, explains Saponara. “Fraud prevention is a very relevant aspect of insurance companies’ strategy, impacting claims rates, prices and customer experience, and needs to be constantly modernized. This is certainly another aspect that led the company to be successful”, he states.
