The financial services industry (FSI) is building a reputation as a leader in using innovative technology to enhance customer experience, drive down costs and combat risk. Artificial intelligence (AI) is a case in point. As a means of understanding and drawing insights more rapidly and accurately from vast amounts of complex data, it’s essential in an environment where that data is pouring in from all sides, all day, every day.
We’ve seen a number of financial institutions harness the power of AI and other types of advanced analytics to help deliver a service to their customers that’s more personalized, compelling and competitive – for example by identifying the best offer to make to an individual customer at the most opportune time, or by enabling customers to get tailored help and guidance online using a chatbot.
There are huge benefits to leveraging AI capabilities in the never-ending fight against fraud and cybercrime, as well as supporting ongoing compliance demands. For example, one insurer ran a pilot project recently using Intel® Saffron™ Cognitive Solution and in just 10 hours detected a fraud ring worth USD 2 million per year.1
Make a big impact behind the scenes
But what about day-to-day operations? Beyond the high-profile customer experience and compliance use cases that everyone immediately thinks of, there are actually hundreds of ways in which AI can make a significant difference behind the scenes.
Empowering employees: By analyzing all the available data about a customer, across all the accounts they have with the bank, and pulling in external information from social media and other public sources, advanced analytics solutions can identify patterns that a busy sales executive would never be able to spot. Now they can get exactly the right customer information, special offers and so on, exactly when they need them, to help them have more relevant, successful conversations with their customers. At the same time, staff can save time on repetitive administrative tasks like checking and filing contracts if an AI-based application is in place to do this legwork for them. For example, one large financial institution has used the Intel® Nervana™ AI solution to review 30,000 documents every day and create summaries that condense the key points into an easily digestible overview for portfolio managers, allowing them to make more accurate decisions faster.2
Increasing efficiency: Across every part of the business, processes and workflows help keep things moving, ensure compliance and deliver on customer expectations. They’re essential but they can also be time-consuming and repetitive, requiring employees to spend time on admin that could otherwise be spent adding value. Bringing automated intelligence into the mix helps take some of this burden off employees, while also ensuring greater consistency and accuracy in the more mundane aspects of office life. Research from Deloitte found that this ‘back office’ application of AI is one of the most impactful use cases in financial organizations today.3
Making smarter ‘things’: Although the FSI industry is a pioneer in the use of many innovative technologies, there are still lessons to be learned from elsewhere, especially when it comes to unleashing the potential of the Internet of Things (IoT). By installing sensors on items around the back office, or even the branch floor, you can begin to automate many maintenance processes, for example, a common practice in manufacturing and automotive companies. Using AI to analyze the data being sent by these sensors, you can then automate and/or optimize the timing of tasks like garbage collection, or replacing printer toner; or ensure electricity is not being used on lighting and HVAC in an empty room. This helps save time and money while maintaining a clean and comfortable working environment.
It's all about your data
The spectrum of potential back-office use cases for AI is broad, and the type of technology you will need depends on your planned use case – for example, natural language processing to help parse and review contracts, or deep learning and cognitive analytics to proactively guide salespeople in their conversations with customers. You can learn more about the different types of analytics in this planning guide.4
However, even the most advanced analytics in the world will only deliver useful results if it has good data to go on. A typical financial services organization may have data in many places: internal databases (often a different one for each business unit) that may or may not talk to each other; customer communications on social media, through the contact center or by email; and external information from government or industry bodies. Pulling this all together into a united view will give you the visibility that you need to achieve the coveted ‘Know Your Customer’ 360-view. It will also provide the foundation you need to be able to develop AI projects as needed over time.
The move from multiple data siloes to a single data ‘lake’ is complex but necessary and therefore it is important to take the time to get it right. Working with your IT team, make sure you’re able to collect and process all the different types of data, and that you have measures in place to ensure that data is governed properly over time and its quality is maintained. Involve your back-office staff early in the process too – the more input they have in planning an AI project that will impact their roles, the more likely they are to support the initiative and see the value. Their input will also help ensure you’re targeting the right use cases and ultimately achieving the best possible results.
Find out more about how to implement your own AI project by reading this recent white paper from Intel.5 6 7
Informacje o produktach i wydajności
Wyniki testów uzyskano przed instalacją najnowszych poprawek oprogramowania i aktualizacji oprogramowania sprzętowego, które służą do likwidacji luk w zabezpieczeniach określanych mianem „Spectre” i „Meltdown”. Po zainstalowaniu tych aktualizacji wyniki te mogą nie mieć zastosowania do posiadanego urządzenia lub systemu.
Oprogramowanie i obciążenia wykorzystane w testach wydajności mogły zostać zoptymalizowane pod kątem wydajnego działania tylko na mikroprocesorach Intel®. Testy wydajności, takie jak SYSmark* i MobileMark*, mierzą wydajność określonych systemów komputerowych, komponentów, oprogramowania, operacji i funkcji. Jakakolwiek zmiana wyżej wymienionych czynników może spowodować uzyskanie innych wyników. Aby wszechstronnie ocenić planowany zakup, w tym wydajność danego produktu w porównaniu z konkurencyjnymi, należy zapoznać się z informacjami z innych źródeł oraz innymi testami wydajności. Więcej informacji można znaleźć na stronie http://www.intel.pl/benchmarks.
Wszystkie podane tu informacje mogą ulec zmianie bez powiadomienia. Aby uzyskać najnowsze dane techniczne i plany rozwoju produktów Intel, należy skontaktować się z przedstawicielem firmy Intel.