Jeremy Jawish, CEO and Co-founder of Shift Technology, shares some insights about how AI improves insurance fraud detection.
By Billy Teo
The Shift Technology story begins with Jeremy Jawish (CEO) and Eric Sibony (CSO) meeting during their studies at the premier French engineering university, Ecole Polytechnique.
The idea of Shift Technology came up during a joint internship at a global insurer, where they were tasked with improving the insurance giant’s fraud detection techniques. Discovering that this area of insurance was ripe for disruption, they dedicated the following months to R&D, ultimately developing expertise in the realm of insurance fraud.
Post-graduation, the stars aligned when Jeremy went on to work on the Fixed Income desk at Goldman Sachs – where he met a top-performing strategist, David Durrleman.
Impressed with his track record, the duo brought Durrleman on as Co-Founder and CTO, solidifying the management team – now complete with a balanced mix of business savvy, mathematical ingenuity, and technical acumen.
Today, the company is serving a growing list of over 70 major insurance companies located in four continents. Singapore is also home to the first incorporated subsidiary of Shift Technology outside of France.
Jeremy had a quick chat with IMpact to share more about the company.
Why did you start up in Singapore?
Many of our early adopter customers in APAC are in Singapore and South East Asia. For this reason alone, establishing Singapore as a hub for SEA made a lot of sense.
Why focus on fraud in the Insurance industry?
Insurance companies are still very traditional and especially in the middle and back office functions, automation and disruptive technology are not yet widely used. The economic damage originating from fraudulent claimants is significant.
Detection solutions are still very much manual and offer room for improvement. Our fraud detection solution, Force, is very advanced and sophisticated and can help insurers to quickly detect suspicious or fraudulent claims.
The benefit for insurers is to be able to pay claims faster, reduce or refuse non-justified insurance claims, and increase the efficiency of the claim process.
What is the conventional way to detect fraud in the industry?
We find that many insurers are still using manual screening processes. Due to the massive amount of data which needs to be screened, these more manual processes are difficult to scale and claim payments can be delayed due to the lengthy process.
Some insurers have implemented rule-based internal models to identify suspicious claims.
However, we found that our AI approach, which is based on real fraud scenarios, yields higher hit rates than these traditional rule-based models.
How does your Force fraud detection technology work?
We have developed several specific algorithms to support the detection process. AI, and supervised and unsupervised machine learning, in combination with tailored fraud scenarios help to detect anomalies in connection with a claim.
Our Force solution will then score these anomalies and push relevant information to the claim and fraud handlers with detailed guidelines and recommendations for the investigation process.
Force provides a decision-making platform specifically designed for insurance fraud handlers to augment their capacity to detect a wide spectrum of fraudulent behaviours, from opportunistic claim exaggerations to organized crime rings.
The idea is not to replace the human fraud handler, but instead replicate his or her deductive reasoning and apply it at large scale. The insurer simply sends us the claims data, and Force takes care of the rest: analyzing the claims, retrieving and cross referencing relevant information from both internal and external databases, and sending scored alerts in real-time.
The solution, which can be rapidly implemented in as little as four months, is designed to seamlessly integrate with an insurer’s existing operational processes, regardless of the technical context.
What gives Force an edge over competing solutions?
I believe we have a strong competitive advantage over its competitors since we specialize only in insurance fraud detection. Additionally, I think we have the most talented AI experts in the field and a differentiated approach to our models.
Lastly, the fraud detection solution is provided in a SaaS-format, which grants the solution much more agility than the products offered by competitors. These factors combine to result in higher conversion rates and better ROI (Return on Investment) for insurers.
Why did Shift Technology pursue a Software-as-a-Service (SaaS) model?
SaaS hosted in industry standard cloud space in the respective country is the most efficient way for Force to operate and process massive amounts of data in a short period of time. Capacity adjustment, data and operations safety and security are other reasons why we operate in a SaaS model. Additionally, our customers benefit from continuous innovation, quicker time-to-market and low costs of entry.
How do you think the AI behind insurance fraud detection will evolve?
AI is already widely used to help with the automation of processes. The ultimate goal of an insurer is to settle justified claims as quickly as possible without compromises on control and governance. Ultimately, the insurance beneficiary will benefit from more stable insurance premiums if fraud can be eliminated.
We believe that in addition to enabling better, more accurate fraud detection, AI is instrumental to enabling the vision of an automated claims process for improved customer satisfaction.
Our next-generation solution, Luke, promises to remove the inefficiencies inherent to claims processing, replacing existing core systems with streamlined automated processes. From the filing of a claim to the policyholder payout; AI will be present every step of the way, reducing end-to-end handling time from weeks to minutes.
How was the experience of being accredited under Accreditation@SGD?
The accreditation process was very detailed. The process helped us to be more aware of the outside-in problems. The questions raised by the Accreditation investigators helped us to formulate our response statements to be more precise and more understandable.
We understand that being accredited serves as a quality and reference label, and helps to foster our credibility in the market in Singapore, and beyond.
How big is the team now, and what are your expansion plans?
We have about 200 employees globally. Singapore is one of our two technology hubs in APAC. The team in Singapore currently has 10 staff, and we expect to expand to up to 20 members soon.
For more information about IMDA's Accreditation@SG:Digital programme, visit https://www.imda.gov.sg/accreditation.
.webp)