Artificial Intelligence (AI) is already proving to be a truly transformative technology and its impact on a whole range of sectors is set to make a significant impression in 2018. In particular, the Financial Services Industry (FSI), often weighed down by legacy IT infrastructure, will reap the benefits of this emerging technology. This evolution was the focus of the recent NextGen Banking London: The AI Revolution event where industry experts gathered for a series of talks and panel discussions on the opportunities and challenges presented by AI in the banking sector.
“AI is the next wave of computing and will enable banks to radically improve efficiency, risk management, customer service and regulatory compliance”
AI will enable banks to radically improve efficiency, risk management, customer service and regulatory compliance. Even today, the technology is helping to augment the work that people do, and soon its applications may well be expanded to include a higher degree of automation and machine-human collaboration. However, to reap the benefits offered by AI, finance firms first need to navigate through a range of technical and cultural challenges.
From customer service chatbots to fraud detection, AI has a wide range of applications in FSI. One area where it is set to make a big difference is individual personalisation. Speaking at the recent NextGen Banking London event, Gayle Sheppard, Vice President and General Manager of the Saffron AI Group at Intel, outlined one case study where AI is already making a difference in this area. The USAA (United States Automobile Association) serves members of the US military and their families, giving them competitive rates on insurance, banking and investment services. The company, which has around 13 million customers, aims to recognise what the next best action is to recommend to each customer, if any action is needed at all.
“What we did was create a personalisation implementation prototype for them, using our advanced AI technology,” explained Sheppard. “We worked with over 1,300 products, using three main data sources including customer interaction history. Using this, we were able to achieve 84% accuracy in sales recommendation.” After being given access to the relevant data, it took the prototype only ten weeks to produce a beneficial result for the USAA. What’s more, this was accomplished using just 1.5 FTEs (Full Time Equivalents) at Intel, and two at USAA, demonstrating the impressive return on investment that AI can provide.
Saffron*, which began as a tech start-up, was first acquired by Intel in 2015, and now Intel® Saffron™ forms a key part of Intel’s AI strategy. Another way in which Intel® Saffron™ AI technology is driving the digital transformation of the banking industry is fraud detection. In one example, Intel helped a customer reduce their fraudulent medical insurance claims; using a combination of one-shot learning, which is based on small amounts of data, and continuous learning which dynamically adapts to larger amounts of new data over time, the technology was able to detect a fraud pattern within just 48 hours. And within only four weeks, it had identified three fraud rings based on the existing data set.
Another key application of AI for FSI is in product testing and defect resolution. “Banks happen to be some of the largest software companies in the world,” explained Sheppard. “One of the challenges we have in software development is that roughly 50% of the issues identified are duplicates. This leads to engineers working on duplicate issues without knowing it. If we can identify those issues that are duplicates, or very similar, and make sure they’re consolidated, we can save time and drive efficiency, reducing the associated costs.”
Intel recently released a suite of artificial intelligence-powered software applications using associative memory learning and reasoning to enable businesses to resolve issues more quickly. Known as Intel® Saffron™ AI Quality and Maintenance Decision Support Suite, it is already used by a number of major firms including global professional services company Accenture*.
“The Accenture Touchless Testing Platform is augmented with artificial intelligence technology from Intel® Saffron™ AI that brings in analytics and visualization capabilities,” said Kishore Durg, senior managing director of Growth and Strategy and Global Testing Services Lead at Accenture, speaking at the time of the platform’s launch. “These support rapid decision-making and help reduce over-engineering efforts that can save anywhere from 30 to 50 per cent of time and effort.”
These examples show some of the key applications for AI in FSI and how they can boost efficiency, detect fraud and enable personalisation of products and services. Banks must take action now to move forward with AI, but the two biggest obstacles are data and culture. There is still lots of work to be done on consolidating data sources and ensuring the quality of data. This is particularly challenging in the banking sector as data tends to be siloed, largely transaction-based and stored on legacy systems. Alongside the consolidation of usable data, banking organisations must also undergo a cultural shift. A willingness to innovate and adopt AI technologies throughout all areas of the business is essential.
“The path to adoption is product,” says Sheppard, adding, “we’ve seen that in every major technology industry. That’s why we’ve made a commitment to introduce products that solve specific problems, that are packaged end-to-end, from data acquisition to natural language processing [NPL].” Products like these based on Intel® Saffron™ technology will help to drive the future of AI in the banking industry, enabling financial institutions to remain competitive in the face of changing regulations and the evolution of Open Banking.
*Other names and brands may be claimed as the property of others
For more information:
Stay connected to technologies, trends, and ideas that are shaping the future of the workplace with the Intel IT Center.