How to unlock value with next-generation intelligent underwriting

Use cases for intelligent automation in underwriting

Insurers are looking for ways to automate repeatable, mundane tasks by overlaying techniques like process mining, robotic process automation (RPA), workflow orchestration, and optical character recognition (OCR) on core administrative systems. Reducing the manual work that results in error and higher human capital spend, allows underwriters to get more done in less time.

Insurers can use intelligent automation for:

• Intake of exposure
• Coverage and risk information from standardized ACORD forms
• Non-standard Statement of Value (SOVs) and loss runs
• Enhanced handwriting recognition using OCR technology
• Ingestion from both paper and digital channels like agency portals
• Comparative raters

Improving underwriting insights

Insurers want technologies that give them better insights into real-time data. Risk profiles change rapidly, and they can’t trust the accuracy of third-party data sources that rely on historical attributes. They need tools to deal with big data and dark web data mining or InsurTechs for exposure analysis, automated risk monitoring and data-driven insights.

Underwriters must look beyond traditional risk analysis methods. Insurers are turning to cognitive risk evaluation methods that consider isolated and accumulated risk profiles, local and regional hazards, and continuous book monitoring. This includes:

• Integration with nontraditional and big data sources for risk augmentation
• Predictive models for submission propensity
• Real-time underwriting appetite for specific exposure classes or insurance to value (ITVs)
• Household-level underwriting evaluation.

Equipping underwriters with the right resources

The role of an underwriter continues to expand and the need for consistency in processes and risk evaluation continues to increase. Insurers are looking to give their underwriters as much information as possible, including pricing and risk assessment aids, just-in-time (JIT) guidelines, up-sell/cross-sell recommendations, and other contextualizing data. It is important for insurers to underwrite the immediate submission and they should also keep the long-term underwriting vision in mind. Insurers should be leveraging intelligent automation capabilities like artificial intelligence (AI) and machine learning (ML) for real-time recommendations; personalized pricing strategies for coverages using predictive analytics; and contextual underwriting guidelines using natural language processing (NLP) — or large language model (LLM) — based chatbots, virtual agents, and conversational AI.

Providing a better view of data to personalize the underwriting experience

Investing in a unified system gives underwriters a true “cockpit view.” Insurers should look for a system that includes an intuitive interface for risk visualization and classification, personalized dashboards with key performance indicators (KPIs) and objectives and key results (OKRs), and the ability to collaborate with internal and external teams using email, chat, chatbots, etc. A better view of data helps insurers provide a personalized experience through the insurance lifecycle.

How intelligent automation is changing the insurance landscape

Accelerating underwriting processes while balancing profitable risk selection is critical to remain competitive. Customers expect quick and easy access to insurance products and services. If underwriting processes are slow, redundant, and complex, brokers may turn to competitors who can provide a faster and more streamlined experience. Improving the process for customers is also crucial because delays can lead to frustration and distrust.

With automation, insurers can increase underwriting productivity and write more profitable business without adding headcount — a necessity for addressing the ongoing talent shortage that will be exacerbated by 50% of the current insurance workforce retiring over the next 15 years.[1]

Insurers have a vast amount of data at their disposal that is often locked in disparate systems, inboxes, and submission documents. A data fabric, which is a component of leading process automation platforms, creates a virtual database to unify all these data sources, letting insurers synthesize information and surface insights in a single pane of glass. This helps insurers to be more informed to achieve predictable outcomes, minimize losses, and increase profitability.

Automation platforms and low code

The future of underwriting will be defined by citizen automation development platforms (CADP), low-code development platforms, and automation across the entire customer intake and underwriting lifecycle.

Low-code/no-code (LCNC) and CADP technologies are becoming more sophisticated and mainstream, and insurers are achieving rapid speed to market and customized automation workflows. Platforms like these, insurers are adopting workflow automation, building web-based apps, bridging data across systems and applications, and generating consolidated visualization across data sources.

CADPs are becoming more intuitive and flexible, and more business units are developing personalized lightweight solutions. Solutions are being developed by the intended user group, more insight and context can be embedded from the start. Additionally, since these platforms enable automation and development in an LCNC construct, business units can reduce their reliance on traditionally high-cost technology talent. They can do more upskilling and increase staff motivation, resulting in improved retention rates.

Opportunity intake and management

Automated ingestion capabilities allows insurers to accept paper submissions and standardize ACORD forms. Even key risk characteristics like SOVs, loss runs, and conditional exposures can be ingested automatically.

With a single underwriting management solution, underwriters can pick up submissions from all intake channels. They can automate repeated processes, such as producer license and registration check, potential account-level Office of Foreign Assets Control violations, and existing account history checks.

Underwriting insights and analytics

AI/ML-based analytical models can predict submission propensity to provide dynamic pricing strategies and cross-sell/upsell suggestions. Insurers can provide cognitive risk insights that enable underwriters to effectively make cross-sales and upsale on submissions.

Underwriters can contextualize underwriting guidelines to specific risks they are evaluating with chatbots powered by NLP/LLM.

The EY-Appian alliance

Together, EY organization and Appian:

• Provide platforms and solutions with the potential to help maximize resources and improve business results, with a focus on financial services.
• Help improve customer experience and achieve operational excellence.
• Support simplifying risk management and regulation compliance for businesses.

The EY–Appian Alliance is applicable to support all functions across the enterprise, such as product innovation, distribution and sales optimization, customer service modernization, finance and risk transformation, enterprise protection and enterprise IT transformation.

The Appian Connected Underwriting solution

Appian Insurance solutions are a new class of packaged and custom software built on the Appian Platform. Customers benefit from faster time to value with packaged applications while maintaining flexibility to adapt and extend those applications using Appian’s low-code design capabilities.

Part of the EY-Appian Alliance, Appian Connected Underwriting is a prebuilt solution that leverages the power and flexibility of the Appian Platform to accelerate underwriting, optimize operations, and improve customer experience by reducing time to quote and quickly integrating disparate data sources into one unified view. With Appian Connected Underwriting, insurers eliminate redundant manual data entry and improve process efficiency using automation tools like RPA, intelligent document processing, AI, workflow automation, and process mining. Appian Connected Underwriting architecture is based on four critical capabilities: