Turn around in almost any city, and you’re likely to see a bank or lender or brokerage on the corner. In fact, in my family’s small town, we have two financial institutions by the same name on either side of a two-lane street. That’s convenience. And, while I love the personal experience I get from visiting my hometown banker, I also appreciate being able to conduct my business after the bankers have gone home to dinner, and knowing that my fraud protection never sleeps.
Financial services institutions (FSIs) of all sizes recognize that they are in fierce competition to deliver differentiated services while meeting stringent regulatory and compliance requirements. Among the earliest adopters of digital transformation, FSIs satisfy these requirements with a range of emerging technologies, including artificial intelligence (AI).
With AI techniques such as machine learning and deep learning, FSIs are capitalizing on massive amounts of data to automate processes, reduce fraud, build closer customer relationships, and meet other customer and stakeholder expectations.
Use case examples
While these institutions are implementing countless AI use cases, these are some of the most typical.
Improving customer service
In a highly competitive industry like financial services, customer satisfaction reigns supreme. To that end, AI-driven applications can help FSIs automate processes to improve customer service while reducing operational expenses.
A few examples:
- AI-based chatbots help customers access information quickly and efficiently, without the need for staff to handle inquiries directly. These chatbots use AI-powered natural language processing and voice analytics to help customers check balances, apply for loans, and access other services.
- With advances in voice-recognition technologies, data analytics, and AI algorithms, natural language processing systems can detect not only the words customers use but also the tone and sentiment behind them. These capabilities help FSIs automate quality-assurance monitoring and work proactively to maintain customer satisfaction.
- Today’s tech‑savvy consumers increasingly seek financial services that are faster, cheaper, and easier to access. With AI on the back end, FSIs are improving their ability to predict consumer needs and preferences and deliver richer, more convenient, and personalized customer experiences—essential to driving revenue growth in a time of rising customer expectations.
Detecting and preventing fraud
For FSIs, AI is now an essential weapon in the battle against fraud. AI algorithms help banks and payment processors spot unusual and potentially fraudulent transactions in real-time so they can immediately stop the activity.
To identify and stop fraudulent transactions, Mastercard leverages machine-learning algorithms running on HPC systems to process large datasets as transactions are taking place. This capability helps Mastercard stop fraud in its tracks without disrupting or delaying legitimate transactions.i
And it’s not just banks and payment processors that are deploying AI. Insurance companies now increasingly use AI to spot suspicious claims by comparing new claims against legitimate claim patterns.
Meeting regulatory and compliance standards
In recent years, FSIs have grappled with a stream of new government regulations related to data protection, customer privacy, risk controls, anti-money laundering, operational processes, and more. AI-driven applications help organizations comply with these stringent regulations by automating the process of identifying, collating, and analyzing data from disparate systems to meet these new compliance requirements.
AI also helps financial institutions automate their “Know Your Customer” requirements. AI does tasks such as verifying customers’ identities, analyzing the types of activities they’re involved in, and evaluating the risk of involvement with money laundering or other criminal activities.
Considerations for capitalizing on AI
Though these use cases are diverse, implementing them poses a common set of challenges, according to my colleague Anurag Juneja, a Dell Technologies senior presales data analytics specialist who focuses on the financial services industry. I turned to Anurag for his thoughts on considerations for FSIs embarking on an AI journey. Here are some of his suggestions for FSIs that want to develop, test and continually refine AI uses cases and algorithms.
The first step in any AI project is to identify the business outcomes and what data the organization has that can contribute to success, Anurag says.
“The use case always drives the infrastructure requirements — processing speed, storage location and tiering, server and storage throughput, and networking bandwidth,” he notes. “The use case also dictates the requirements for speed, data and application protection, recovery time objectives and recovery point objectives, and so on.”
Infrastructure is a critical piece of the puzzle for organizations building and deploying AI. Anurag notes that AI-driven applications require the power of massively parallel processing systems to manage and analyze petabytes of data, as well as fast interconnects and high-performance storage. AI development also requires advanced solutions for data protection, data security, and data privacy, along with intuitive tools for managing the compute, storage, and networking infrastructure.
To capitalize more fully on AI, financial services companies need to integrate data from diverse sources — including enterprise systems, transaction processing systems, social media, and more, Anurag notes.
“This data integration enables better processes and services, such as gaining a 360-degree view of the customer to personalize offers,” he says. “You can’t get there if your data lives in disconnected silos.”
Real-time data processing
In a world of AI-driven applications, many use cases require real-time processing of data as it streams into private and public cloud environments, as opposed to traditional batch processing. Anurag says that this shift to streaming data requires mastering new software technologies that enable the immediate processing and analysis of data.
Open-source training models
According to Anurag, organizations can save time and effort with open-source AI training models that help jumpstart AI development efforts. Some of these models can be accessed and used via public clouds, while others are designed for hybrid environments that leverage a combination of on-premises and public/private cloud resources.
“Some FSIs leverage public cloud by uploading data from sources where hardware infrastructure is not viable, train the machine learning model using that data in the cloud and finally deploy it to on-premises data centers,” he says. “This approach helps to keep costs low and performance high using HPC-backed infrastructures. What’s more, this hybrid model by itself has opened up channels to explore data points that traditionally were not possible to access otherwise.”
To train AI models and operate AI-driven applications, FSIs need someone who understands both AI algorithms and the hardware infrastructure that powers the training and use of the AI models, Anurag notes. The silos established in IT and data science communities have led to gaps that must be shortened for a successful AI initiative.
“Data scientists typically do not understand enough about how the infrastructure design plays a critical role in driving machine learning and deep learning models,” he says. “Every model requires its own set of network bandwidth, IOPS, CPUs, and/or GPUs to be successful and performant enough to drive the application layer goals. Hence, bridging the gap between IT and data science is imperative to delivering a successful AI application.”
With rising customer expectations, new regulations, new technologies, and market disruptors around every corner, financial services institutions need to embrace digital transformation. AI is an essential element in digital transformation, and in the coming years, it will become all the more critical to the success of FSIs.
To learn more
For a deeper dive into the topics explored here, see the Dell Technologies “Connected Finance” eBook. This handbook examines the disruptive trends sweeping through the modern financial landscape and highlights how the Internet of Things, advanced analytics, artificial intelligence, blockchain, and virtual reality help drive business differentiation and value.