Insurance is all about managing risk, but with AI it only gets better. Let’s take the example of an oil and gas company. It produces terabytes of operational data daily– both structured and unstructured. Here, insurance companies can connect the data to predictive analytics systems to anticipate levels of degradation, perform automatic defect inspections, predict potential failure rates, operational risks, etc. 

(Source: NVIDIA)

In another example, one of the insurance companies developed an AI/ machine learning algorithm for predicting the likelihood of flooding in the area using historical and geospatial data and inputs from digitised documents. This allowed the company to model a potential market with 83 percent accuracy, reduce throughput time in underwriting by 10x, and improve case acceptance by 25 per cent. 

Thanks to the adoption of AI, NLP and analytics, the insurance space is undergoing massive digital transformation across the value chain, including underwriting, policy servicing, customer support, and claim settlement. As a result, companies are investing heavily in digitising claims and processing platforms more than ever. 

As AI and NLP enable digitisation, the power of analytics boosts the effectiveness of extracted content by leveraging predictive modelling. Insurance companies are using these cutting-edge technologies to conduct processes faster and effectively. With the increased availability of data, more and more insurance companies are using AI to eliminate redundant manual tasks. 

Underwriting helps insurance companies analyse a company’s financial performance. However, with the rising competition in the space, the function is no longer limited to underwriters.

According to the GlobalData Engineering Technology Trends Survey, 62 percent of insurers said they are investing in AI, and nearly half of them believe it will be critical to business development in the coming years. Thus, it becomes important for insurance companies to leverage disruptive technologies and transform underwriting. 

In this article, we will look into some of the use cases insurance companies can pick to increase AI, NLP, and analytics adoption within the underwriting space. 

Here’s how you can increase AI, NLP and analytics adoption

In any insurance organisation, submission, followed by quote and bind, is crucial. Unfortunately, this process can stretch to a few months, resulting in high underwriting costs. 

Effective and efficient application of AI and analytics can reduce processing cycle time to a few days and elevate the customer experience. If we refer to the previous article and leverage the prioritisation framework, the use case of the new business submissions will fall under the top priority bucket. It requires interpretations from varied unstructured sources, including emails, applications, quotes, proposals, contracts, etc., leveraged in the underwriting process. 

Depending upon the type of risk, the new submissions set-up process requires manual keying in of 10-50 fields from multiple documents, such as received date, insured name and address, broker information, policy inception date, no number of layers, insured value, type of risk, etc. 

The enablement of submissions ‘AI data pipeline’ will ensure a robust foundation that can be repurposed in the digitising underwriting process. This will also support digitising other documents, such as line-slip/binders/declarations/market reform contracts, statement of values, loss-runs, invoicing, survey reports, and others. 

Another use case is the digitisation of line-slips, standard contracting documents in the London market. They specify contract terms of risk-sharing amongst different insurers. The semi-structured sections of line-slips contain risk-driven parameters like assured name, written lines, type of risk, premium and limited, brokerage, risk clauses, etc. This digitisation will enable faster quotations from underwriters. 

Role of analytics in improving digitisation outcomes 

The benefits of digitisation are limited if we just extract the information and do not enrich the data. The extracted data can be outdated or partial, requiring the underwriter’s team to validate such information, like industry risk types, broker codes, names, etc. The enrichment of data not just delivers cost savings by manual interventions but also increases the effectiveness of decision models. 

Data enrichment enables better throughput of risk prioritisation and triaging analytics models, essential in any end-to-end process optimisation. The predictive model’s application ensures faster new case allocation to relevant underwriting teams and case prioritisation, eventually faster quote processing. The power of analytics just does not end here; it can further support pricing by forecasting the potential nature of risks.  

For optimised outcomes, companies need to adopt AI solutions such as: 

• Computer vision and NLP techniques to make extraction possible 
• Predictive modelling will enable case prioritisation and better pricing 
• Workflow application which orchestrates all components 
• Deep domain knowledge 
• Human-in-the-loop to assess overall throughputs before being sent to realisation 

Wrapping up

Application of AI and analytics in underwriting improves risk capturing, decisioning, and customer experience. Moreover, companies should adopt a use-case prioritisation framework followed by flexible, scalable, and domain-centric AI solutions.

This article is written by a member of the AIM Leaders Council. AIM Leaders Council is an invitation-only forum of senior executives in the Data Science and Analytics industry. To check if you are eligible for a membership, please fill the form here.