Insurers are adopting digital technologies at a pace never seen before. But manual underwriting processes for extracting data continue to cause delays across the insurance value chain. P&C insurers are under tremendous market pressures in the after-effects of the COVID-19 pandemic and rising expectations from policyholders for a better digital insurance experience. As an industry, insurance is witnessing a growing number of production deployments and plans to scale various AI and Hyperautomation initiatives. Companies that fail to make strategic investments in AI today will struggle to compete and stay relevant in the next few years.
Underwriting is at the heart of insurance. Assessing risks to set profitable and competitive prices in an increasingly low-margin marketplace is a difficult one. Only a limited number of highly skilled experts are qualified to complete this type of work. But many of the tasks within the underwriting process are repetitive and could easily be automated. For example, the commercial insurance submission processes are highly manual. Specialists spend a significant amount of time sourcing and consolidating data before performing the risk analysis. While migration to remote work introduced newer working methods during the COVID-19 pandemic, it also revealed the longstanding issues with manual, paper-based processes. Most insurers are heavily dependent on paper-based processes of varying degrees. Brokers and underwriters can no longer count on passing paper around the office or turning to a coworker sitting next to them for a missing piece of information. Streamlining activities like straight-through processing and continuous underwriting require uninterrupted real-time data access to assess complex risks quickly. However, much of this data is trapped in digital insurance documents of different types and formats, and almost 80% of it is unstructured. The capacity of skilled underwriters is underutilized on such routine tasks and data enrichments that bots can do better, faster, and with more accuracy. With AI-powered automation and human-in loop, insurers can free up underwriting teams to focus on customers and growth.
Underwriters spend much of their time in their email inbox, accessing and responding to new applications, prequalifying questions, questions seeking additional information, and so on. Underwriting a new submission is a manual process that sometimes takes underwriters days or even weeks to complete. Here is how traditionally, new business submissions were processed.
- Broker emails underwriter to submit a new opportunity.
- Pre-qualification of the opportunity is performed by the underwriter
- Broker submits ACORD applications, supplementals, and other supporting documents for formal underwriter review through posts, faxes, or emails
- Underwriter performs an initial assessment for risk and additional information requirement
- Underwriter confirms coverage quotes matching submitted applications
- Quote proposals are bundled and emailed to the broker
- If customer approval is received, quote is bound, and underwriter begins the issuance process
Here’s what the new submission intake process could look like with embedded Hyperautomation, AI/ML, and IDP:
- Broker submits a new submission through digital channels like emails and web portals through to be reviewed by the insurance company
- Insurance company’s AI software uses Intelligent Document Processing (IDP) to analyze the submissions, extract and classify the contents.
- IDP helps to collect data necessary to underwrite the submission, while NLP and ML techniques are used to validate the content for business rules and missing or incomplete data
- The collected data is published in downstream applications where underwriters review the information and make final evaluation related to acceptability
- Coverage quotes are created and reviewed by AI/ML models to ensure accuracy
- Quote proposals are sent to the broker
- If customer approval is received, broker requests policy issuance
The ability to quickly and comprehensively analyze the commercial risk submissions and offer multiple quote proposals for different participation terms for which an insurer is comfortable is a daunting task. Insurers need to utilize hyperautomation and AI to supplement and expedite these processes to remain competitive.
When insurers spend all their time manually extracting data and reviewing spreadsheets, they miss opportunities that move more quickly and offer excellent options to their brokers and customers. With Kanverse’s Insurance Document Processing, underwriters can spend less time reviewing websites, applications, and contracts, process submissions in hours instead of weeks, and focus on complex decision making and relationship building which ultimately leads to gains of new written premium and decreased claim activity.
Powered by a proprietary multi-stage engine that combines computer vision and natural language processing, Kanverse automates submission intake across multiple channels, classifies the types and variations of the documents, performs data extraction with industry-leading 99.5% accuracy, validates the data based on business rules and files it in Insurance software systems.
Contact Kanverse.ai for more details.
About the Author
Kingshuk Ghosh, Product Manager, Kanverse.ai