Fraud detection is at the center stage for the insurance industry. Regulatory compliances and business disputes leave the door open for a number of frauds to happen. Better management of fraud risk and compliance exposure is critical for the insurance industry to innovate and grow. Better fraud management ensures that there is no opportunity for a conﬂict to arise. Digital Foresight is one such forward-looking, proactive solution that helps organizations to detect and materially crush fraud.
Careful, Looks can be Deceiving
The insurance industry is confronted by an unprecedented number of experienced, well-funded, focused and motivated adversaries with the intent to steal focus from the insurers. Posing a much greater threat than the old fraudsters, the domestic and, now, “transnational” bad actors further add to the problems. The funds stolen directly and indirectly decrease proﬁts, threaten solvency, and reduce the number of customers—leading to negative ratings and regulatory action.
This paper presents groundbreaking modernization technologies designed to reduce fraud and enhance risk management capabilities. While these solutions have particular application in the property and casualty segment, they also serve many other disciplines and operations of an insurer.
Fraud Threats or Red Flags
Transnationals are organized crime or terrorist groups functioning as rogue states or cartel operatives. They are often sophisticated groups with substantial funding, access to technical skills, and employ advanced technology as their “weapon” of choice. They have the ability to operate eﬃciently like any guerilla force moving quietly, ﬂuidly, professionally and successfully. The systematic theft and fraud perpetrated by transnationals has the potential to quickly bring an insurer to its knees. These groups are far more organized than the “organized crime” of yesteryears.
Understanding the transnational bad actor is a prerequisite for creating an eﬀective solution to safeguard the integrity of an insurer’s claims operation.
The transnational threat to domestic and global insurers is extreme across all lines of business, and represents an enormous economic loss to the insurance industry. Insurance giants are not always the preferred targets. In fact, smart transnationals may ﬁnd softer targets in tier two insurers. They employ smart people with resources and budgets to focus on the insurance company's automatic payment thresholds and operate under the radar.
The transnational bad actor’s motive is not just to “steal money.” Their modus operandi is to launder cash, monetize ill-gotten or illegal goods, and also steal money. They may select hard targets or soft targets, and balance their strategy and “market.” The question is how to eﬀectively combat transnationals at the lowest net cost with the highest impact and with solutions that can morph as quickly as transnational bad actors deftly change their strategies. Periodic audits also sometimes fail to reveal behavioral patterns that could lead to detection and apprehension.
Real-time Fraud Detection: A Must-have for Insurers
Insurance companies are looking to identify and detect potential fraud in real-time at the point of sale (POS). It helps in identifying potential losses from overpayments, high claims frequency and fraudulent or professional claimants. As the criminals’ ability to beat the current system is increasing, the insurer’s claims processing system must include far more nimble, smart, and anticipatory platforms, tools, and protocols honed by experienced professionals.
Why Use Real-time, Always-on Systems for Fraud Detection?
Real-time solutions safeguard insurer’s capital, markets, and reputation today; solutions that are ﬂuid, ﬂexible, and quickly adaptable to changing transnational strategies.
There are many aspects of counter fraud that present opportunities for reliable science-enabled systems and processes to stop material losses posed by the transnational. Financial intelligence-driven systems and practices can thwart criminals and simultaneously fuel a broad range of actionable outcomes. They serve underwriting, actuary, regulatory, Special Investigations Unit (SIU), marketing, loss control, and the overall enterprise. These real-time, always-on systems also complement Know Your Customer (KYC) information enablement.
Coforge’ subject matter experts, highly specialized tools and protocols, and organizational integrity oﬀer insurers the best defense against claims-based fraud losses.
Fraud: An Intractable Problem
By looking back at the insurance industry claims terrain and how it was traversed, we have a layered understanding and awareness of what needs to be done today to combat fraud.
Hampered by claims systems built in the 1980s as well as high operating and loss adjustment expenses, carriers, third-party administrators, brokers and risk managers face intractable challenges as they seek to modernize claims processes, including:
- Legacy claims systems that are becoming less ﬂexible, adaptive, and more diﬃcult to replace as SMEs retire
- Transactional data that is error-prone and not well maintained. A recent survey of 30 CIOs from large carriers rated data quality as 3 (low) on a scale of 1 to 10
In addition to modernizing claims processes, companies with legacy systems must improve user experience and provide on-demand capabilities to maintain brand loyalty with agents and customers.
Know Better, Act Faster
To modernize claims systems and combat fraud eﬀectively, carriers must know when the fraud is going to happen, and who its participants are. Emerging technologies that incorporate new sources of data and advanced analytics can help companies identify speciﬁc risks and players to predict and mitigate fraud.
If the idea is to attain clarity in the claims operating environment, the ﬁrst step is to eliminate noise from data. In other words, the carrier’s internal data must be cleaned. Companies will then be in a position to employ optimized information and intelligence strategies and tools and resources that go far beyond data mining techniques.
Addressing data warehouse noise helps organizations gain operational eﬃciency and increase understanding of counter fraud tactics —preventing the likely leakages. It is more of a barn burning, and then getting the idea out of the horse.
To gain accurate, actionable, outcome-based intelligence, we must adopt ways and means to reduce noise levels in the data warehouse content.
This is done by real-time fusing of all relevant internal and external information. By harvesting external data from all sources that are related and inferred to the subject points in the data warehouse, we ensure that the people, places, and things are constantly reﬁned in real-time, and access to external data is achieved.
By adopting this approach, we quickly see that the fog in the data warehouse battle space has lifted. The ability to discern the actual patterns of operational behavior, people, and service providers among other things, is clearer now and more actionable than imagined.
Achieving Predictability within the Existing Technology Framework
Noise in the claims system, claims data, and claims external mission space allows transnationals to operate within the claims environment without getting caught.
With our proactive approach, we have the ability to generate forward-looking, high ﬁdelity, and actionable outcome-based factors for predictability. However, to attain true predictability in fraud detection, the solution must present actionable data to individuals that need it.
Modernization does not necessarily require new systems, high expense, or major investments in the claims environment. Modernization enables predictable outcomes leading to a material reduction in fraud-related losses, and even plain leakage. This can be achieved by deploying advanced tools and techniques in claims operations that work within the existing technology framework.
Need for Experience-based Solutions
The market abounds with fraud prevention ideas, solutions and vendors. The vendors must have real-world operating experience combined with real-world technology expertise to implement successful solutions.
It is diﬃcult to imagine how a vendor can deliver pragmatic, meaningful fraud-free solutions without deep experience in actual insurer operations. After all, understanding the balance between how a risk is chosen, priced, sold, administered, and kept is a must to decide whether a claim is valid or not. The ability to build, deliver, and keep a solution requires real understanding of the insurance enterprise.
An experience-based solution optimizes the organization’s ability to walk the ﬁne line between the CIO and the executives driving a modernization project in the insurer enterprise. During the resource deployment discussions, organizations can decide to increase the comfort levels with the solution timing and structure with technology functioning as an integral component of the business, leading to long-term success.
Add Digital ForesightSM to Your Artillery to Combat Insurance Fraud
By optimizing data to quickly identify and predict fraud, Digital Foresight represents the ability to detect, and materially crush fraud.
The word prediction has a jaundiced interpretation. Nevertheless, predictive in an actuarial sense yields conﬁdence. Fraud can be anticipated with conﬁdence based on a host of variables. However, few are comfortable with predictive as a certainty.
When one avenue is closed, transnationals with their bright and capable technologists remain in the game by ﬁnding another path into the claims system. They use the insurer’s claims engine to steal, launder, and repurpose the insurer’s money.
A solution to predict and mitigate fraud outcomes must include a balance of fraud-crushing activities that collectively anticipate with conﬁdence where, how, when and by whom a fraud will be committed. Achieving this agility in claims management will close all the new avenues at the earliest.
Predicting and mitigating fraud with conﬁdence does not require reformation of the claims system. It is not a Capex idea, and can be delivered as an add-on to the existing claims operation regardless of the platform.
Predicting When and Where the Fraud Will Occur
Digital Foresight from Coforge is designed to provide insurance companies with real-time analytics to improve their business operations across the enterprise. Digital Foresight (at point-of-sale or claims’ ﬁrst notice of loss) fuses transactional information with big data to predict when and where fraud is going to happen rather than where they have already happened.
To battle fraud, Digital Foresight captures numerous classes of data that constantly ﬂow into the marketplace, and information about the claimant, distribution point or anyone who touches the transaction. It is not a data warehouse but data, information, and intelligence on everything relevant to the transaction both inside and outside the organization.
The solution incorporates the most advanced form of geospatial correlation (identifying all people, places, and things that are linked to a claimant, bill, provider, and ultimate claim). Everything is considered—the claimants’ misdeeds, where they live and work, who they talk to, and who they owe money to. In short, the entire web of circumstance that causes a person to act good and bad is considered.
Digital Foresight fuses relevant internal and external data that is directly associated with an instance or, more importantly, linked by strong inference, and applies pattern analytics. It is a discipline that employs artiﬁcial intelligence techniques to ﬁnd patterns in the data.
Consider nodes of people, places, and things that are related to a bad actor’s activities. Also, include the potential of internal collusion with the insurer. All the data is linked, valued and developed into patterns of behavior, or more importantly, a signature of behavior trajectory. From this analysis, emerges the predictive outcome.
While examining the concept more closely, we have come to a conclusion that organizations have data that they can buy to develop correlations and ideas. However, the experience of applying signature and behavior science belongs to only a few.
Signature analytics is a demonstrated skill of the Digital Foresight team who pioneered this science and its application over the past 10 years. Signature analytics, in short, means developing an understanding of the patterns that the people follow to conduct their lives, their business, and the world they traverse. Sometimes referred to as human terrain, signature analytics allows the claims operations intelligence function to discern the behavior patterns of all related parties and quickly recognize when those behaviors are leading to fraud, or tell the claims operation which nodes of the claims ecosystem need attention to strengthen the customer engagement.
A Discrete Geographic Perspective
The Coforge team not only provides this level of behavior analytics but also reduces the claims fraud wilderness to the most discrete geographic point of understanding—critical to achieve fraud prevention through predictive analytics.
Consider a granular data resolution perspective of 1 km x 1 km; a hexagon that connects people, places, and things. Claims operational intelligence utilizing Digital Foresight applies graph-based assimilation of data relating to people, places and things viewed from a discrete geographic perspective. Using data graphing ontology, Digital Foresight wraps pattern recognition in a vector hexagon of geography at a 1 km x 1 km scale.
Digital Foresight enables the insurer to see the most ﬁnite links, nodes, and relationships between people, locations, and events (a loss incident, provider services, relationship between people, or other events). The claims organization is empowered to see data linkages more accurately in real-time.
The produced data used across the enterprise brings high ﬁdelity. The most important characteristic is currency. Currency is used in monetary value, time, and reduced losses. The Digital Foresight approach enables the insurer’s data and the insurer’s claims operation to acquire currency—a winning attribute. Currency is all about the data being immediately actionable, usable, latest and available in real-time.
Digital Foresight continuously acquires the latest data matching relevant points of business with everything that the claimant is doing today. It is not a data warehouse but data, information, and intelligence on everything relevant to the transaction both inside and outside the organization. It looks at all the role players including agents, claims representatives, witnesses, lawyers, providers, and all the known bad guys to discern patterns of fraudulent activity. The results are then fed into a real-time claims system.
Digital Foresight applies highly advanced, agent-based science that identiﬁes the earliest indication of a fraud at the most discrete level of identity (literally at the level of a bill from a disparate service provider and so on). This deliverable is a highly ﬁnite and discrete tuple (ordered list) of disruption to conventional pattern analytics. The beneﬁts of the process are:
- the ability to see the earliest indication of fraud much faster than it was previously possible
- the ability to predict the next emergence of fraud with high ﬁdelity at the earliest using data relating to people, places, and things in the entire claims ecosystem
When a pattern is spotted with an agent, distribution channel or customer, the ﬂagged account may be forwarded to a SIU working with law enforcement to build a case for indictment and conviction.
Claims that must be processed for compliance are monitored electronically for post-claim payment fraud. Continuous review of social media and other data can identify that a workers comp disability claimant is not currently disabled and payments must be stopped to avoid fraud.
Digital Foresight is a powerful tool to appropriately and quickly pay claims, evaluate distribution channels, and work more eﬀectively with stakeholders. A 4-5 month implementation timeline enhances its appeal.
The availability of open domain information analyzed by Digital Foresight eliminates the need for vast data stores. The data is harvested every hour of the day based on keyword ontology—a technique to establish links between the data elements using relationships, words and events.
The data comes from various sources including social media. Aggregators can harvest, slice and dice data from Facebook, LinkedIn, and Twitter inexpensively. However, these three sources account for only six-tenths of the one percent of the data available.
The primary expense is to employ data scientists and engineers to develop and optimize solutions in the client’s environment. These professionals create algorithms to collect and assemble intelligence using their understanding of data and social sciences.
Putting Intelligence into Action
Some property and casualty companies are implementing computerized (sometimes manual) adjudication processes for selected lines of business (termed fast-track claims adjudication by some) while promoting employees to handle more complex bills and overarching claims. However, increased automation introduces opportunities to game the system, and expands opportunities for newer types of fraud.
Early Identiﬁcation and Mitigation
Coforge has the capacity to implement a fraud identiﬁcation and mitigation solution to address the “variable auto-adjudication engine” problem. Hosted on the cloud and added to an existing claims system, a variable auto-adjudication solution can be an important part of the modernization project.
Variable auto-adjudication also applies to what is considered to be the best optimized treatment of data, especially when fusing disparate structured and unstructured data. The data assimilation involves graphing, which speeds up the process and allows data to be better aligned and readily linked to related and inferred data—increasing understanding.
Impossible to Game
The solution transforms the fraud threshold from a straight line to something like an ECG. Through advanced logic, the threshold constantly resets at variable intervals by variable amounts, making the engine impossible to game. This solution potentially saves an enormous amount of money as criminals seek easier targets.
For example: If the auto adjudication threshold is $X,XXX for Y coverage loss, but ﬁts the behavior pattern for fraud, then through Digital Foresight, the insurer can either deny the claim or evaluate in-depth before paying.
Objections might be raised based on the potential degree of diﬃculty, and the time and cost involved. These factors are indeed variable but the solution is highly feasible considering the science involved and the experience of the team.
As an illustration, consider a request from the United States intelligence community to map the entire globe on a discrete hexagon framework for all the relevant data points identiﬁed. The solution would be cloud-based for the best data and fastest data provisioning. With experience and the necessary data science, the task can be accomplished in minutes without any real technology deployment.
Game-Changing Fraud Prevention Technique
Digital Foresight eﬀectively utilizes GPS to enable early fraud identiﬁcation and loss prevention in the run-up to a declared weather-related catastrophe or the aftermath of a disaster.
The technology can also be used to:
- Inform Risk Management: Catastrophe modeling using Digital Foresight helps identify risks with advance disaster preparation and notiﬁcation—ensuring protection of property and people
- Determine Proximate Cause: Basis catastrophe modeling, it gets easier for insurance companies to ﬁgure out whether wind or rain is the proximate cause of damage to a building. This allows for a sound judgment on accepting or rejecting claims and claims payouts.
Coforge has developed a catastrophic underwriting modeling solution, “Exact+,” that performs catastrophic underwriting and claims handling of storm, earthquake, terrorism, climate change and other scenarios, and supports underwriting and claims processes. Once the solution is built, the technology can be employed to scan the entire universe of data and address business opportunities.
Managing the Face of Fraud Tomorrow
Digital Foresight is a forward-looking, proactive solution that addresses the industry’s intractable problems and enhances business opportunities.
With unique real-world, practiced skill, the distinguished Digital Foresight team at Coforge is delivering real results for various clients, such as:
- Fraud Reduction by more than 10%: Fraud reduction from the Digital Foresight and variable auto-adjudication solutions is not novel. Savings have exceeded 2% in the ﬁrst full operational month and can easily rise to 10% per year. Shutting down an organized crime group that was hitting one company on their casualty book demonstrates the true power of this technology
- Increased Eﬃciency: In addition to reducing fraud, variable auto-adjudication makes claims processing more eﬃcient, reducing the cost of processing a claim. Relevant data captured during the claims process can also be exported to showcase omni-channel capabilities to stakeholders on their mobile devices
- Improved Eﬀectiveness: Modernized systems will help carriers work more eﬀectively with customers, agents, and employers, improving customer satisfaction while reducing IT cost. Improved data analytics will increase stakeholder performance as companies drive more business to the top performing lawyers, contractors, body shops, etc.
- Increased Competitiveness: Fraud reduction will free up resources for other initiatives. Agents and customers will choose carriers based on their customer experience, time to process claims and time-to-market of new products and services
Coforge leads the eﬀort to modernize claims systems with Digital Foresight and our approach to claims auto-adjudication. As regulatory changes in monitoring Own Risk and Solvency Assessment (ORSA) takes eﬀect, interest in these solutions will also accelerate.
Deﬁning the New Normal
The explosion of data from telematics and IoT-connected devices will have to be managed in real time to execute processes eﬃciently. Companies are experimenting now with property telematics—placing devices to measure environmental factors or earthquake movement at home, to track cargo, and to record personal, bus, and police car driver behavior. All this data will have an insurance risk management and premium impact.
With Digital Foresight, Coforge is in a good position to manage your data. We understand people, behavior trajectory, analytics, and human terrain. Our Digital Foresight team includes some of the top global experts who have nuanced understanding of industry dynamics.
The Coforge Thought Board:
Safeguarding Your Reputation
The Bottom Line: New Capabilities to Fight Fraud
Tier one insurance companies are beginning to see greater value in tier two consultants basis their experience with primary consulting ﬁrms. In addition to agility and excellent core competency, clients can get more attention for the dollars they spend by establishing strategic partnership with Coforge.
Companies are putting increased emphasis on a business solution as opposed to just an IT solution. Competitors do an excellent job by implementing IT solutions based on the client’s requirements. However, Coforge’ approach is to help clients develop solutions based on their requirements and select, build, test and deploy the best solution. We also provide other services including consulting, infrastructure, business process outsourcing, operational and performance analytics, and a framework for fraud detection.
Coforge works with a variety of clients including carriers, brokers, third-party administrators, and self-insured employers. We partner with other product companies, establish alliances or develop frameworks or, if appropriate, use our own products to deliver business value to clients.
By focusing on the insurance industry and a few other verticals, Coforge has been successful in giving operating experience to our employees and delivering high quality solutions—diﬃcult to ﬁnd in any other IT vendor.