Insurer-case-study

Summary

Our client, one of the UK’s largest P&C insurance companies, offers consumers and businesses a wide range of insurance products, including motor, home, business, travel, and more. They faced challenges in their call center operations, with long hold times and errors affecting efficiency.

To address this, they partnered with Fractal to implement FractalGPT – GenAI Knowledge Assist. This AI-driven solution improved knowledge retrieval and response generation, leading to smoother operations and better customer interactions. 

Challenge

The insurer faced challenges due to long hold times in its contact center, stemming from the difficulties faced by the agents in locating relevant information related to policies and procedures. 

 This resulted in response delays, frequent supervisor interventions, and inaccurate responses. 

  • A large workforce of over 500 agents managing high volume of calls exceeding 300,000 per month
  • Approximately 30% of calls requiring escalation to supervisors. 
  • About 10% of interactions resulting in errors leading to potential financial damages. 

Solution

To solve these problems, the insurer partnered with Fractal, a leading provider of B2B AI solutions. 

Fractal implemented its FractalGPT – GenAI Knowledge Assist application. FractalGPT is an AI-powered solution designed to revolutionize knowledge retrieval within large enterprises using the power of large language models and generative AI. 

In eight weeks, FractalGPT- GenAI Knowledge Assist was deployed and customized on the client’s AWS environment, where it was integrated with the internal SharePoint and policy repositories and several other backend tools. The virtual assistant was further customized by training it on thousands of documents covering various topics such as policy FAQs, claims, premium payment, policy closure and more.  

It leveraged a large language model (LLM) to perform semantic search and response generation, ensuring that the answers were relevant and consistent with the source documents. The approach ensured that no confidential or company-specific information left the company tenant. 

Knowledge Assist, built on Amazon EKS, Amazon Bedrock, and Amazon Open Search Service, ensures efficient and accurate information retrieval without extensive training. It seamlessly handles complex documents, including those containing tables, across multiple file formats such as PDFs, Docs, and PPTs. Additionally, the solution integrates with OCR to extract information from images. 

Result

Large P&C insurer optimizes its call center operations using GenAI on AWS

Reduction in AHT

Streamlined interactions, faster resolutions

Large P&C insurer optimizes its call center operations using GenAI on AWS

Reduced supervisor escalations

Agents empowered to handle queries independently 

Large P&C insurer optimizes its call center operations using GenAI on AWS

Decrease in erroneous response

Accuracy soared

Large P&C insurer optimizes its call center operations using GenAI on AWS

Answer accuracy

Boosted customer satisfaction

Large P&C insurer optimizes its call center operations using GenAI on AWS

Projected annual savings

A substantial impact on the bottom line

FractalGPT solution is now available on the AWS Marketplace; click here to learn more.

Insurer-case-study

Business challenge

Businesses worldwide are collecting, analyzing, and using data to streamline operations, better understand their A leading property and casualty (P&C) insurer wanted to modernize its policy and claims systems, aiming to tailor offerings across different brands. The company leveraged a legacy system, which hindered its ability to:

icon1

Launch New Products Quickly

Respond to changing customer needs and market trends.

Modernizing the policy and claims systems for a large P&C insurer

Improve User Experience

Enhance quote-to-policy conversion ratio by addressing current digital limitations.

Modernizing the policy and claims systems for a large P&C insurer

Capitalize on Market Opportunities

Customize and innovate products for usage-based, embedded, excess, and retention insurance.

Modernizing the policy and claims systems for a large P&C insurer

Optimize Pricing and Underwriting

Enhance data quality, reduce fraud, and enable seamless online self-service.

icon5

Create a Data Lake

Facilitate analytics, machine learning, and better underwriting outcomes by consolidating fragmented data infrastructure.

To overcome these challenges, they were looking to migrate from the legacy system to Guidewire, a comprehensive software suite designed for the insurance industry, to streamline core operations such as policy administration, claims management, and billing processes.

Approach / Solution

The P&C insurer needed a holistic strategy to tackle the challenges posed by their outdated legacy systems. They worked with Fractal to construct a highly scalable and robust system, leveraging a range of AWS services. This comprehensive solution unfolded in several key stages.

  1. Building the system
    • The system’s foundation was established by creating a secure storage space for data called a data lake and data stores, facilitating advanced analytics and machine learning capabilities.
  2. Moving the data
    • The historical data was moved from the old legacy system to the new one, leveraging AWS services like AWS DMS and Glue ETL.
    • Data accuracy and consistency were maintained throughout the migration journey with the help of Change Data Capture (CDC) using AWS DMS.
  3. Going digital
    • The successful data migration took place by transitioning from S3 to intermediate RDS storage, generating PDFs for policy notes, and handling special characters with the help of FPDF and ANSI encoding via AWS Glue.
    • This resulted in the migration of policies to the new Guidewire 10 environment, enabling the seamless transfer of comprehensive information encompassing renewal details, home policy specifics, and general notes about each property.

Results

Modernizing the policy and claims systems for a large P&C insurer

Launch new products faster and meet customer and market needs

Modernizing the policy and claims systems for a large P&C insurer

Improved user experience with higher policy sales

Modernizing the policy and claims systems for a large P&C insurer
×
Match individuals across data sources without a unique key

The Big Picture

A leading credit bureau provided analytics and intelligence support to local credit rating agencies. The company had access to multiple data sources, such as bank data, voter IDs, and tax returns, from where it pulled information for individuals and created their credit score.

However, it was not possible to use information from all these data sources as there was no single unique key. For example, in a bank data set, an individual would have a driver’s license number, but in a voter ID data set, the same person would be identified with a voter ID number. The company couldn’t know that the individual was the same in both data sets. Therefore, there was a need to create a logic to match these different data sets without having a unique key.

This meant that the company needed an improved framework of working with unstructured addresses data so that it could provide a single view of customers across different data sources. It needed to drive measurable performance improvements by improving match accuracy and reducing false positives.

Transformative Solution

To solve the company’s challenges, a solution was deployed to match datasets using names and addresses, since this information was present in all data sources. Since the format of names and addresses were different everywhere, the solution needed to create intelligent and fuzzy logics to standardize names and addresses for mapping purposes.

The approach took raw data and deployed a name and address matching algorithm that was configurable at different levels. The solution incorporated a search capability along with optimization and improvement of matching. Three key steps were:

  • Data standardization: Data was cleaned and normalized to remove components not adding value to addresses. Addresses were segregated into logical components: house number, locality information, and pin code.
  • Address search: The approach searched the request address into candidate data using pin code (and derivatives) as a key.
  • Name and address matching: This step used Fractal’s dCrypt to match all the addresses in a key value pair with request address and selected top 100. For top 100 addresses, the corresponding names were also matched and the best output was generated on the basis of name and address matching scores.

The final output provided a list of names and addresses from the candidate data which match the name and addresses from the reference data. Using the algorithm, a matching score was generated between two strings which could be compared with a base matching score already present in the client’s sample file. Randomly selected samples were manually checked and a confusion matrix was created for both algorithms.

The Change

As a result of the engagement, the company achieved several benefits:

  • Improvement in accuracy by 10% on 11 million household addresses.
  • Incorporation of a search capability in the matching algorithm.
  • Three different algorithms were used for matching as opposed to a single algorithm for name and address matching, which led to better coverage and efficiency.