By Franck Coison, Industry Solution Director, Atos
In a recent PwC survey, 52% of those in financial services said they’re currently making ‘substantial investments’ in AI and 72% of business decision-makers believe that AI will be the business advantage of the future. But first, let me first define what we mean by AI. While there are many forms, we consider here four main types: voice and facial recognition; natural language processing; machine learning; and deep learning. These can be used in various domains through chatbots, document analysis, process automation, or predictive analysis.
In financial services, robotic process automation (RPA) is increasingly common. This technology is perfect for automating relatively simple, repetitive tasks. In contrast, AI can be used to automate more complex tasks that require cognitive, or ‘intelligent’, processes. This kind of intelligent automation is now in high demand. While RPA is appropriate for back-office and accounting processes, when it is combined with AI, any process including customer facing activities can be automated.
Given this potential, there are a wide variety of uses for AI across financial services.
- Customer services. This is one of the most common applications of AI in financial services. Instead of client service executives having to work through hundreds of emails manually, AI can ingest the emails, understand their meaning, and prepare an appropriate answer that the client executive can check and submit with one click.
- Sales and customer intelligence. Again, a fast-growing area, AI is deployed to gather and analyse customer data and intelligence to give business development teams new insights, sales leads and recommendations for the ‘best next action’ to develop the relationship and drive forward a sale.
- Operations. Intelligent automation is a potent way to drive efficiencies and improvements across end-to-end processes, such as insurance claims processes.
- IT services. AI can pinpoint whether an application or piece of hardware is likely to fail, massively increasing effectiveness and resilience of IT infrastructures.
- Fraud prevention. AI is increasingly critical to effective fraud management, by detecting and eliminating fraudulent payments or claims.
- Cyber security. As cyber threats grow and get more sophisticated, AI can be used for predictive analytics that can detect cyber attacks, even before they happen.
Despite what we may see in the news, AI does not, and should not, replace human beings. There are two dimensions to this: AI is here to augment, rather than replace human beings, as human supervision is needed to ensure that AI algorithms deliver the expected results; and secondly, AI is still in the learning curve and will not do everything after day one.
Many of the stand-out benefits of AI are around customer satisfaction. For example, with AI, if we interact with a company online (usually via the website), our questions get answers that are instantaneous, accurate and relevant to our situation. This is what many of today’s customers, especially millennials, want. And there are other major benefits. Quality and accuracy improves significantly by ruling out the potential for human error (again enhancing customer satisfaction and service). And, of course, there are cost savings. If you can use AI, you can increase efficiency and productivity and reallocate your workforce to higher value roles.
Making AI a success
While some financial services are building centres of excellence in AI, many are still exploring the benefits, looking at how to accelerate delivery and identifying what the technology can do – and what it can’t do. Whatever the maturity level in terms of AI, there are a couple of lessons to be learned:
Focus on business pain points
As with any new digital disruptor, it’s important to focus on what you want to achieve rather than the technology. Building a team of AI experts and then asking them to deliver value can make it extremely difficult. Instead, starting with a real understanding of the business and the right pain-point, such as ‘I have a customer satisfaction problem in this area’, then using AI to solve the problem will prove the benefits of AI and gain more traction.
It’s also important to manage internal expectations: AI is not about replacing the human brain.
Another recipe for success is to consolidate all the intelligence or ontologies in the same place in the organisation, rather than spreading it across all entities. This will accelerate the industrialisation of AI, as knowledge will be capitalised and the scope of use cases that can be managed will expand. AI is rising up the agenda for financial services because of its huge impact both on customer satisfaction and operations, as well as its fast return on investment. In two or three years’ time, expected return will at last be evidenced at company level, but for companies who delay – will it be too late?