It’s no longer rare for a financial institution to have deployed artificial intelligence technology in at least some aspect, whether it be customer-facing chatbots, or back-end uses like risk and compliance and fraud monitoring.
Yet many banks and credit unions still struggle to implement AI broadly, or lack a coherent long-term strategy for its use. An Accenture report noted that while three quarters of banking executives believe AI and machine learning will transform the industry, they also report many challenges related to its use. Chief among them are reskilling employees, effectively deploying the technology across the enterprise, and seeing how AI will generate new revenue streams.
One requirement to effectively using AI technology is to start small and identify pilot scenarios that can grow in complexity, KeyBank’s Dean Kontul told The Financial Brand. Kontul is Chief Information Officer for Voice and Chat Automation and Contact Center Delivery.
Services that scale with you.
Achieve a better return on your marketing investment. Leverage behavioral data and analytics to target the right customers with the best possible offers.
Early Results Show Large Savings
KeyBank is one of the largest regional banks in the U.S. With assets of $182 billion, it stands solidly between the megabanks and community financial institutions. While it has a large tech budget, the regional doesn’t have the near limitless budget a large global bank has. That’s why Key has taken a “test and learn” approach with its AI deployment, specifically around conversational AI.
“Key started its conversational AI journey a couple years ago by establishing an Enterprise Center of Excellence for AI and Chat Capabilities,” Kontul explains.
It was out of this center that the bank developed its conversational assistant, known as MyKey, which launched September 2022. The product was developed using Google’s conversational AI technology and was tested first with internal employees before rolling out to customers.
Test and Learn:
Before rolling out customer-facing AI tools, Key hired additional analytics and product people and first tested the technology internally.
Partnering with Google enabled the bank to build a secure, scalable conversational AI solution that could be used to support all areas of the bank. Initially it is being used to improve contact center efficiency and customer experience. “We hired a dedicated team of analytics, product and technology people, as well conversational design resources,” says Kontul.
Currently, MyKey listens to conversations in real time, responds with financial knowledge content and summarizes client concerns, according to Kontul. That reduces contact agent training times while improving agent productivity. In the coming months, the bank plans to add improvements by continuing to leverage Google machine learning capability, he adds.
In the first month after the AI initiative went live, Kontul reports there were nearly 3,000 daily sessions with MyKey, for a total of 250,000 interactions. The digital assistant is also reducing cost to serve by solving customer problems within the app rather than using other channels with an 84% “containment rate” so far.
Kontul adds that Key will continually monitor the performance of the service and make upgrades and changes accordingly.
Silo Alignment Is Necessary
Kontul believes it’s crucial banks commit to company-wide efforts to digitize — on the front end, but especially on the back end — with strict alignment between the CEO and the CIO and IT staff. For Key, that meant creating an enterprise strategy for managing virtual assistants coupled with data collection and analytics to judge how effectively they perform and then making tweaks as necessary.
Banks and credit unions also need to figure out how AI can overcome the naturally siloed arrangements of functions that typically exist across financial institutions. Connecting these siloed systems often involves relying on manually intensive human intervention, according to a report from the Economist Intelligence Unit.
“For example, loan origination has typically been done by a front office team in one part of the bank, using highly classified information about businesses requiring a corporate loan,” the EIU report reads. “Yet the credit risk decision on that loan may rest with a separate team using a different system containing the relevant data.”
- How Community Banks Can Use AI to Improve Sales and Marketing
- How Data, AI & Machine Learning Supercharge Personalization for Banks
- The Strategy Behind Truist’s New AI-Powered Digital Assistant
- The Complex Relationship Between Chatbots and Bank Call Centers
The Need to Re-Skill Employees
Given the need for such manual intervention, it means employees not familiar with AI or without a technical background may need to learn additional skills. For KeyBank, addressing this started with the creation of its Future Ready Initiative in 2018. Part of the initiative included hiring employees familiar with AI technology but also creating a “culture of continuous learning” to upskill existing employees.
Building a culture of continuous learning requires creating the time to learn, grow and practice new skills; technology platforms to enable employees to own their careers; and training to support development in both hard and soft skills, Kontul explains.
KeyBank created an internal learning center designed to give employees new skills and potentially lead them on new career paths.
“To that end, we’ve created internal platforms and invested in external resources to help employees learn how their unique experience, expertise and skills align with new opportunities,” says Kontul. “We emphasize that each employee is the CEO of their own career and on a journey of their choosing to explore and prepare for new opportunities. Team members looking to build or enhance their skills have access to Udemy for Business, an online learning platform providing 5,000+ courses in business and soft skills.”
For example, through this program, the bank may seek to transition workers that already have deep operational experience into new product analyst and build roles, Kontul states.
Ultimately, KeyBank aims to have “the most engaged workforce” and foster a culture of “ambition and continuous learning.”
“This creates non-traditional career path opportunities,” says Kontul, “and leverages the experience with new technology.”