TL;DR:
- Nordic bank DNB has achieved significant success with conversational AI-powered virtual assistants.
- The virtual agent Juno supports customer service departments, answering over two million inquiries in 2022.
- Juno has become a crucial tool for customer agents, providing support and advisory assistance.
- The virtual agent suite developed in collaboration with boost.ai enhances efficiency and accuracy.
- Demand for conversational AI is growing as users and customers appreciate its benefits.
- boost.ai is exploring the use of large language models to improve content creation and training.
- DNB Bank’s virtual agents, including Aino, Hugo, Fix, and Justina, cater to different areas of the organization.
Main AI News:
The impressive success of DNB Bank’s conversational AI-powered virtual assistants over the past three years has been nothing short of remarkable. Among these digital agents, one stands out prominently—Juno. With a staggering two million inquiries addressed in 2022 alone, Juno has been instrumental in supporting the bank’s customer service departments, catering to approximately 1,200 users on a daily basis.
DNB Bank has embraced the power of conversational AI, employing a suite of five virtual agents across various customer and employee-facing scenarios. While Juno may be an internal virtual agent, it has evolved into a highly intricate and invaluable application within the organization, as stated by Jan Thomas Lerstein, the head of emerging technology at DNB Bank.
Juno acts as an indispensable support and advisory tool for all customer agents, empowering them to provide exceptional service to customers. Since its launch in 2020, Juno has played a vital role in assisting customer service agents and advisers by streamlining their access to essential routines and facilitating the retrieval of crucial information. As a result, Juno has significantly expedited response times for handling customer inquiries.
In 2022 alone, Juno furnished valid answers to a remarkable 83% of agents, effortlessly tackling an average of seven questions per user daily. Jan Thomas Lerstein expresses his satisfaction with Juno’s progress, noting that its accuracy is consistently improving. However, he also acknowledges that the lower accuracy rate in some instances is often attributed to questions falling outside the service’s scope or relevance, particularly for new agents unfamiliar with the tool’s intricacies.
DNB Bank embarked on a fruitful partnership with boost.ai, a Norwegian provider of conversational AI software, to develop these virtual agents. Juno leverages boost.ai’s platform, which incorporates advanced filtering functionality. This filtering capability enables Juno to provide tailored responses specific to individual departments, allowing agents to receive accurate and contextually appropriate information. Furthermore, the feedback function within boost.ai’s platform plays a pivotal role in enhancing the virtual agent’s performance and capabilities, facilitating its continual development.
In the past year, there has been a noticeable surge in demand from organizations seeking novel applications and innovative problem-solving approaches through conversational AI. Henry Vaage Iversen, the CCO and co-founder of boost.ai, affirms that users and customers not only embrace this automation but also desire more of it due to the evident benefits it brings to both work life and the overall customer experience.
Looking to the future, boost.ai aims to leverage open and large language models to enhance content creation and refinement. Generative models have proven to be invaluable for creating test data and training resources, significantly simplifying the task for AI trainers and further augmenting their efficacy.
Juno, DNB Bank’s virtual agent, has an extensive repertoire, being well-versed in over 3,400 topics, each tailored to suit the unique requirements of seven different areas spanning corporate banking and the private market. According to Mr. Lerstein, DNB Bank’s exploration of AI dates back to 2017 when they first experimented with automation and improvements in customer-facing services. This journey led to the creation of a suite of virtual agents based on the boost.ai platform.
Apart from Juno, DNB Bank boasts a roster of other virtual agents designed to cater to diverse needs. Noteworthy among them is Aino, a customer-facing virtual agent that has impressively automated over 50% of incoming chat traffic within six months of its introduction, engaging with more than a million customers. Additionally, the bank utilizes Hugo, which assists employees with HR-related queries, and Fix, a virtual agent handling inquiries directed to DNB Bank’s IT service desk. Furthermore, the organization has enlisted Justina, a virtual agent specializing in answering legal questions to provide comprehensive support to its employees.
Conclusion:
DNB Bank’s success in harnessing conversational AI agents demonstrates the immense value and potential of this technology in the market. By leveraging virtual agents like Juno, the bank has witnessed significant improvements in customer service, efficiency, and employee support. The growing demand for conversational AI signifies a shift towards automation and enhanced customer experiences. As technology continues to evolve, businesses across various sectors should consider adopting conversational AI to gain a competitive edge and meet the rising expectations of their customers.
