Insurance Product Cost Optimization

Estimates from McKinsey indicate that if AI is implemented fully in the insurance industry, a potential annual value of up to $1.1 trillion can be achieved.

Although insurance is an old industry and remains highly regulated, insurance companies are recognizing AI-driven solutions that can augment their technological capabilities so that their business can become leaner, faster, and more secure. AI has the potential to transform the insurance experience for customers from frustrating and bureaucratic to fast, on-demand, and affordable. Insurers are waking up to the idea of using cutting-edge AI to find, harvest, and analyze data from both the surface and deep web held across millions of academic papers, patents, government reports, databases, journals,  and news items. It will be used to find signals and generate trends that can help businesses make important decisions about the future.

Insurers use historical data to study their customer base and forecast their exposure to claims. Improving the breadth and depth of the variables that drive this process can enable insurers to adjust the pricing of their products and premiums. This enables them to attract more customers, factor in a customer’s specific needs instead of a generic, one-size-fits-all approach. Customers might soon expect flexible insurance such as on-demand pay-as-you-go insurance and premiums that automatically adjust in response to accidents, customer health, etc.

AI has already started to change the experience in the insurance industry with predictive analytics. By applying advanced analytical techniques enabled by machine learning, accurate forecasting, a key output of business intelligence, has become possible while conducting in-depth studies around the vast quantum of data held by insurers. Hence, intelligent data analysis can incorporate a wider range of risk factors to price risk and policies more accurately.

Non-health insurance fraud in the US, says the FBI, is estimated at over $40 billion per year, which can cost families between $400–700 per year in extra premiums. By employing neural networks and machine learning, fraud patterns can be recognized and used to improve insurance companies’ risks and actuarial models. This can potentially lead to more profitable products. Chatbots using neural networks can be utilized to answer the bulk of customer queries over email, chat and phone calls and understand them. This can free up significant time and resources for insurers, which they can deploy towards more profitable activities.


Insurers face a multitude of challenges in today’s unforgiving economic climate: low interest rates, stagnant revenue growth, greater price transparency, and customer cost consciousness, sweeping regulatory changes, and lack of growth are just some of these challenges that have impacted the profitability of insurers. Many insurance companies are now competing on price, and there are some who can’t sustain this competition. Improving costs and optimizing profits is imperative to address challenges and drive up revenues to enable reinvestment in growth. Along with these legacy claims and underwriting systems, rigid operating models hinder the adoption of new business processes.

Something else that affects profitability is the requirements/loss ratio. If $50 is paid in claims for each $100 collected as premium, then its loss ratio is 50% with a profit ratio/gross margin of 50% or $50.  Insurers can collect premiums more than the amount paid in claims. Conversely, insurers that consistently experience high loss ratios may be in bad financial health. They may not be collecting enough premium to pay claims, expenses and still make a reasonable profit. For example, in the late 1990s, loss ratios for health insurance (known as the medical loss ratio, or MLR) ranged from 60% to 110% (40% profits to 10% losses). As of 2015, the average US medical loss ratio for private insurers was 91.8%. Following the Patient Protection and Affordable Care Act, a minimum MLRs of 85% for the large group market and 80% for the individual and small group markets is now mandatory, and insurers who do not spend 80-85% of their premiums in health care costs must issue rebates to consumers.

Another major challenge is that with the sudden advent of different advanced analytic methods, new techniques and methods are creating a higher barrier to entry. Traditionally, machine learning has shown itself as an uncharted domain — unfamiliar, difficult to explain, and complicated to implement. Since specialization is needed, the involved processes mandate that actuaries and quantitative analysts learn new methods in automated statistical analysis and programming techniques from almost scratch. Insurance companies have struggled to build in-house data science organizations. Coupled with scarce resources and slow technical infrastructure, insurers face a real challenge while onboarding required talent and implementing solutions offered by machine learning.

Solution Approach platform provides out-of-the-box development frameworks.  The project was started with the relevant environments, which were then created automatically. Development images configured based on pre-defined templates were installed on-premises or in a development VM within the infrastructure. This enabled authentication using LDAP, seamless project setup using Bitbucket, Jenkins, and Docker (ensuring build and deployment without software compatibility issues). 

The platform made available by leverages the latest ML and DL tools while preparing models. It includes Pachyderm-based data versioning, Kubernetes, Kubeflow, and Spark-based ML and DL. It also includes an Istio-based service mesh-enabled microservice architecture, and ELK-based monitoring capability, contributing to a reduction in latency’s MLOps platform allows establishing high-end Alluxio and Presto-based efficient data connectivity and collecting data from diverse sources.

The platform can potentially be fed with input data in the form of insurance databases that include size, sales channels, product mix or geography of an insurer (often viewed as cost drivers) and customer data including available social media exposure, details of actual medical conditions reflective of the customer’s actual experience instead of a general overview.

The details collected were added as exploratory variables by using libraries and analyzed. The attributes obtained were used for categorization (employing Pachyderm-based data versioning) and then performing univariate, bi-variate, and Bag of Words analysis — for both structured and unstructured datasets through xpresso Exploratory Data Analysis (Data and Statistical Analysis).  Different datasets and their different versions were easily controlled and stored into an xpresso Data Model (XDM)-enabled data store that enabled easy retrieval and storage of datasets/ files into internal XDM. This was achieved using two excellent features of

  1. Data Connectivity Marketplace libraries
  2. Data Versioning

The xpresso Data Pipeline Management (Rapid Model Training and Experimentation) uses Kubeflow-enabled pipelines. Multiple experiments using different models and datasets could be created, tested, paused, and restarted to gain better insight.

The inference gathered was used to provide the required analytics, which in turn provided insurers the knowledge to negotiate any increase in deductibles and retention limits. It also enabled deriving optimal premiums that are tailored to meet a customer’s specific needs based on their historical data than a conservative estimate, and thus optimize MLRs and profits.

How can help Finance Organizations transform their journey to cognitive AI solutions is an AI/ML Application Lifecycle Management Platform. enables complete lifecycle management of AI/ML solutions, addressing the AI transformation journey of enterprises on any cloud platform of choice. offers functionality essential for building AI/ML solutions – primarily enabling data scientists to rapidly build predictive and prescriptive models. The platform provides a user-friendly interface to develop, deploy, and manage AI/ML solutions at scale. In addition, supports the incorporation of these solutions into business processes, surrounding infrastructure, products and applications.

Key benefits of include:

  • Empowers data scientists to transform AI/ML research into solutions 
  • Improves the productivity of data scientists by enabling them to focus on the business problem, developing algorithms and rapid experimentation of models 
  • Addresses the shortage of skilled data science resources with automated workflows, toolkits and frameworks 
  • Manages AI transformation journey costs without any wastage of R&D efforts  
  • Provides an enterprise-ready and secure environment for complete lifecycle management of AI/ML applications
  • Enables at-scale deployment of enterprise AI/ML applications on-premise, cloud (AWS, GCP, Azure), or hybrid environments

Additional details on can be found at: We can schedule a demo of the platform for anyone interested in learning more.

Have Any Questions?

Need more information about the platform?