Two Key Technologies Driving Machine Learning In Financial Services

Andy Hirst

Predictive powers

Many people wish they could predict what will happen next in the world, but most predictions are quickly assigned to the wastebin of time. That’s because, as hindsight shows, many unforeseen factors change the prediction models.

However, in a narrow domain such as financial services, predictive analytics can be very effective when combined with current technology. Consider black box technology, which enables insurance companies to analyse large data sets. Add data on weather, road conditions, and driver claim histories and demographics, and predictive models can quickly be created to determine key connections. But because predictive models degrade over time, it is critical to continue refining the model with fresh data, and to use machine learning techniques to make and improve accurate predictions.

Analysing text

Another challenge in the era of Big Data is the task of reading and summarizing huge amounts of unstructured data in emails, blogs, and articles. Let’s look at the example of an insurer of an oil tanker moving past the Somalia coast. The underwriter does not have time to read every local report, news article, or email on topics that may affect the tanker, but text analytics allows the underwriter to extract semantic information and sentiment from these sources in a structured form that can be easily searched and understood. A risk or sentiment score can then be assigned that identifies critical information around insured assets and allows an employee to focus on critical risks rather than “information noise.”

Combine text analytics with natural-language chatbots, and you have the ability to make simple inferences and answer questions about the unstructured data. These models are constantly refined over time to improve accuracy. Some analysts, such as Mckinsey, predict that the use of intelligent chatbots could render many call centers obsolete within a decade.

However, predictive and text analytics can create even more value when inserted in business process applications that use transaction data to create intelligent applications. SAP provides new wizards to quickly create predictive models and automation tools to train these models on data over time, adding business value. For instance, when embedded in claims management systems, premiums can be quickly altered to reflect the true risks and aligned revenues with exposures.

Expect greater use of predictive analytics with machine-learning capabilities in customer, risk, finance, and claims applications in the coming years, along with the ability to embed these technologies into existing applications. Natural-language processing and text analytics of unstructured data will also boost the productivity of customer service and knowledge management roles over the next decade.

Learn how to bring new technologies and services together to power digital transformation by downloading “The Future Services Sector: Connected Services for Continuous Delivery.”

This article originally appeared on Finextra.


Andy Hirst

About Andy Hirst

Andy Hirst is vice president of Banking Solutions, SAP Banking Industry Business Unit, at SAP. He is responsible for driving the success of the SAP go-to-market strategy in Line of Business Cloud Applications and Analytics in Financial Services. Previously, Andy was responsible for Capital Markets solutions for banking. Andy is an expert in Big Data and analytics use cases in financial services and has been involved in many digital banking initiatives for banks.