Expert Guidance: Opportunities for Automation in the Insurance Industry
October 10, 2019
By Radhika Seth Pande, Senior Solutions Consultant, WorkFusion
The potential for automation in the insurance industry is limitless. Across different geographies and sometimes even within the same company, different processes are in different stages of digitization. But where to start an insurance automation program? What are the areas most suited for digital transformation, and where can processes be automated end-to-end? Where will robotic process automation in insurance deliver sufficient results, versus where intelligent process automation in insurance would be more desirable?
Let’s dive in.
This key workflow is an important part of Know Your Customer (KYC) programs. The customer submits identity and income documents along with the standard application form. An insurance analyst performs pre-underwriting activities which include document checking and verification — i.e., reconciling supporting documents against the information offered in the insurance application form. Next comes data entry of the application details into the target system, which for large, established insurance companies is often a mainframe application.
A combination of optical character recognition (OCR), machine learning and robotic process automation (RPA) can help with information extraction from forms (structured data) and supporting documents (unstructured data) and speed up the process of entering such details into the customer database.
Underwriting includes rules-based assessment and scoring of each insurance application:
Risk (health) — example: Mortality charges, and hence, premiums, will go up in cases where an applicant is a smoker, especially when weighed against their current age.
Financial limits — example: If the net worth of the applicant today is $X, their insurance coverage (death benefit) cannot exceed, say, $10X.
Creditworthiness — example: What is the credit rating of the applicant, as per external agencies such as Experian or TransUnion CIBIL Limited?
Duplicate policies — example: Does the applicant already have a policy in their name?
Robotic process automation can handle all of these tasks. A rules engine makes it easy to create and maintain rules-based operations. (Some options to consider are Drools or OpenL Tablets business rule management systems.) Alternatively, RPA tools can also be used to create and manage underwriting rules as well, provided the right kind of governance framework and trained resources are available. Systemic checks or web-scraping (navigating through websites and other sources to fetch relevant information) also can be handled by RPA.
At this stage, the pre-underwriting checks have been done and the underwriting decision has already been made. The premium amount and coverage have already been decided and need to be communicated to the customer. RPA can be used to draft standard emails like this. Once the customer accepts, other steps including payment instruction and processing, generation of the policy document and final issuance of the policy, all can be automated using RPA.
Existing policy holders can submit requests such as address change, update of bank mandate, policy top-ups, policy cancellations, etc. Processing for each of these requests is amenable to automation. Machine learning can be applied to the step of classifying emails and RPA can be applied toward performing the required steps and updates at the back end.
At the time of claims submission, a wide range of activities have to be performed, and each is a great candidate for process automation. These include a multi-way match of the claim form against the supporting claim documents and images — such as medical certificates and reports (in case of life insurance or health insurance claims); photos of damaged baggage plus flight boarding passes (in case of a travel loss claim); police reports, driver’s licences and vehicle damage photographs (in case of an auto claim).
In a nutshell, a combination of web scraping for fetching information from websites, computer vision for images, OCR plus ML for document images, and RPA can help simplify the work of an insurance and/or claims analyst where bots play the role of maker and people are checkers. Process automation creates a complete audit trail of each step performed. Customer experiences are improved as the back office functions occur in a timely manner with fewer errors and delays and less annoying back-and-forth communication with customers.
Automation is the lever that can help transform your processes, radically elevate your employee satisfaction and improve your customer journeys. Automation is a path to delighting your employees and your customers!
About the author: Radhika is an automation solution evangelist with a keen understanding of banking, financial services and insurance processes.