What Industrialising Machine Learning Means For South Africa’s Insurance Sector

The world as we know it continues to change exponentially, and at pace. Macro-economic headwinds, rising inflation, supply chain disruption and changing consumer behaviour have come together to create a perfect storm for businesses across all industries. For insurers, it has meant a shift from providing products and services to delivering exceptional customer experience.

Technology lies at the heart of the ability to provide customers with the experience they have grown to expect. We now live in the era where there is an increased focus on user experience – which has led to hyper-personalisation in products and services, as well as engagement, interaction, and customer support. Companies are expected to find out exactly what their clients want and need – and respond.

Emerging technologies, like Artificial Intelligence (AI) and its subsets including Machine Learning (ML), offer the potential to enhance the end-user experience, automate administrative processes, provide deep analytic insights, optimise workflows, and reduce costs.

Research by McKinsey indicates that there are a number of new technologies that are driving these and other benefits across sectors – and one of the key emerging trends creating value and positive impact is industrialising ML.

Industrialising ML workflows are the software and hardware solutions, systems and processes that bring AI and ML into production for real-world business use. The systems accelerate the development and deployment of ML and support performance monitoring, stability, and ongoing improvement. ML tools have the ability to help businesses shift from pilot projects to viable business products, resolve modelling failures during production, and overcome limits around capacity and productivity.

Harnessing the benefits of industrialising ML

Organisations that can successfully industrialise ML can shorten the production time for ML applications from proof of concept to product by 90 percent and cut down on development resources by up to 40 percent – ultimately helping drive efficiencies, reduce cost and enable organisations to make smarter, better-informed decisions. Recognition of its potential drove a $5 billion investment in industrialising ML globally in 2021.

Use cases and the ability to drive positive impact using industrialising ML differs from industry to industry: supporting the development of new drugs in pharmaceuticals, for instance, and enabling key services such as risk management and fraud detection in the financial services sector. Fraud detection is one of the three main areas where the technology can create value in insurance, along with pricing and reserving.

Reserves are the funds that need to be set aside by an insurance company for future claims it may have to pay out. Maintaining sufficient reserves is critical to ensure that insurers are able to pay future claims, and to preserve solvency if claims made are larger than predicted.

Fluctuations in reserves impact the bottom line, but it can be difficult and time-consuming to predict the reserve levels required. Industrialising ML can help with the development of more accurate reserving models to optimise results and improve liquidity and the overall bottom line. This in turn improves planning and decision-making around reserves.

Industrialising ML can additionally streamline and enhance pricing models – allowing insurers to move away from generalised linear pricing models where customers are grouped according to factors such as age, gender, location or car model they drive to individualised, dynamic pricing.

Creating a competitive advantage

Machine learning can increase accuracy and reduce volatility to optimise pricing based on individual risk factors. Industrialising ML also allows insurers to be agile, flexible and respond more rapidly from a pricing perspective – acting as a competitive advantage and differentiator in an ever-changing landscape.

Utilising machine learning technology can also serve as a competitive advantage in the detection and prevention of fraudulent claims. Having the correct processes, pipelines and data sets in place will allow insurers to detect and act on fraudulent claims, driving down costs and reducing turnaround times on claims. Industrialising ML helps reduce workloads that would before have taken hours to complete, to minutes or even seconds, enabling workers to focus on more important tasks.

More and more, insurance companies are investing in the technologies – like industrialising ML – that boost the customer experience by improving efficiencies and reducing costs. They are also increasingly partnering with insurtechs to develop and implement relevant technology-enabled solutions. LAUNCHPAD, for instance, is Guardrisk’s insurtech initiative, and intends to partner with entrepreneurs and venture capital investors to develop solutions that utilise technologies like industrialising ML in order to address specific business and industry challenges as well as changing customer needs.

Industrialising ML is still in its early stages in South Africa. Although insurers, technology companies and other businesses across sectors will have to overcome challenges with setting up the correct processes, pipelines, and systems to unlock the potential and value of the technology, all indications point to an unparalleled ability to meet customer needs and improve their experience while simultaneously improving business efficiencies and cutting costs.

• Xolani Nxanga, Managing Executive at Guardrisk