Leading Norwegian bank streamlines second-line support with over 1,200 daily active users
DNB, Norway’s largest bank, is on the cutting edge of using conversational AI to improve various organizational functions. Since 2017, the bank has worked with boost.ai to implement a comprehensive conversational AI strategy that consists of no less than five virtual agents operating across customer- and employee-facing use cases.
Its customer service virtual agent, Aino, automates over 50% of all incoming chat traffic and is the primary channel for customers visiting DNB’s website. The bank has also applied this ‘chat-first’ approach internally with virtual agents like Hugo that assists employees with HR-related queries 24/7, and Fix, the IT support bot, that handles incoming inquiries to DNB’s service desk.
In its efforts to continue to deliver the best possible customer service experience, DNB also identified that conversational AI could be used as an effective tool to address some challenges in the workflows of its human agents and banking advisors.
DNB's results at a glance...
- 1,200 daily active users
- 80,000 conversations per month
- 2 million inquiries in 2022
Document management woes
DNB's customer service agents and banking advisors were facing challenges in providing quick and accurate assistance to customers through channels such as phone, email, and live chat.
To assist customers, agents need to access various routines - processes outlining how they complete a procedure or answer a query - that comply with the bank’s guidelines and maintain consistency across the organization.
The previous method of using a document management system was found to be inefficient and time-consuming for the agents. The process required them to manually search to find the relevant routines, which often led to delays in providing assistance and negatively impacted the customer experience.
The need for a more efficient solution that could give agents quicker access to the information they required was evident, and DNB began to explore the use of conversational AI to address this challenge.
One virtual agent across seven business units
In March 2020, DNB launched the virtual agent Juno to assist its agents and advisors in quickly and easily accessing the various routines they need to follow when assisting customers, in addition to other key information.
"Juno has been a game-changer for our customer service agents. It has made it a lot easier to find information, making their jobs easier and customers receive faster and more accurate responses to their inquiries."
- Maia Sognefest, Juno Product Manager, DNB
Juno uses the boost.ai platform’s filtering functionality that allows a single virtual agent to provide answers across multiple business units, without the need for each unit to have its own standalone bot. Filtering enables Juno to tailor its replies so that agents each receive a unique, department-specific response, instead of a generic one.
This functionality has allowed DNB to significantly scale Juno’s capacity to answer questions and has resulted in the virtual agent replacing the bank’s previous data management system in its entirety. Agents now exclusively use Juno to access routines specific to their departments more quickly and effectively than ever.
Juno covers 7 different business units across Private and Corporate Banking, like Everyday Banking and Loans
Juno can answer questions on more than 3,400 topics, with each of these then further customized towards seven different areas within the bank, split between Corporate Banking and the Private Market.
Juno’s popularity among the agents that use it has resulted in other departments within the bank requesting to have access to the bot, with additional business units scheduled to be onboarded in the future.
"Juno has been a game-changer for our customer service agents," said Maia Sognefest, a Product Manager at DNB who leads the Juno team. "It has made it a lot easier to find information, making their jobs easier and customers receive faster and more accurate responses to their inquiries.”
Additionally, the boost.ai platform integrates seamlessly with a number of APIs and robotic process automation (RPA) systems used by DNB to further streamline agent workflows and eliminate the need for them to switch between systems.
"With Juno, we have created a feature-rich conversational interface for our customer service agents, providing much more than just a traditional chat service"
- Jan Thomas Lerstein, Head of Emerging Technology, DNB
Juno can automatically check the status of different systems within the bank and inform agents of any issues or disruptions so that they can better assist customers.
Upon opening the chatbot, agents are greeted with a welcome message that informs them of the status of the bank’s various services and any recent changes made to routines. This helps agents to provide accurate and reliable information to customers and improves the overall customer experience.
A feature-rich conversational interface for customer service agents
The implementation of Juno has led to a significant streamlining of DNB’s customer service operations. With around 1,200 daily active users, Juno answered over 2 million questions in 2022 alone, demonstrating its high level of usage among agents who see the bot as not just an essential tool, but as a helpful digital colleague that makes their lives easier.
Jan Thomas Lerstein, Head of Emerging Technology at DNB, was impressed with the results. "With Juno, we have created a feature-rich conversational interface for our customer service agents, providing much more than just a traditional chat service," he said. "Juno has become an application of its own.”
Juno handles an average of 80,000 conversations per month, providing a wealth of data for Sognefest’s team to analyze. Among these conversations, some have been observed to last several hours, with agents leaving the chat window open throughout the day to easily access Juno whenever they need assistance. By utilizing the boost.ai admin panel, the AI trainers who maintain Juno are able to mine this data, identifying areas that require updates and ensuring continuous improvement of the chatbot's performance.
“We fully appreciate the guidance provided by boost.ai throughout Juno’s implementation,” adds Sognefest. “As well as the full control we have over the chatbot’s content via the easy-to-use admin panel.”
The next step for Sognefest and her team is making Juno more proactive. This will include having the virtual agent send preemptive messages to customer service agents regarding changes in routines or the bank's systems, enabling them to stay informed and adhere to the latest policies.
With Juno and DNB’s other virtual agents contributing to a holistic approach to conversational AI, the bank continues to lead the way in using boost.ai's technology to streamline processes and increase the efficiency of its customer service operations.