Íslandsbanki's AI-Powered Virtual Agent Automates 50% of Chat Traffic in Six Months
- Infrastructure as a Service (IaaS) - Virtual Private Cloud
- Platform as a Service (PaaS) - Application Development Platforms
- Finance & Insurance
- Retail
- Sales & Marketing
- Chatbots
- Onsite Human Safety Management
- System Integration
Íslandsbanki, one of the three major banks in Iceland, was facing a challenge in managing its customer service. The bank was trying to make the banking experience more digital and less branch-heavy, while simultaneously improving its infrastructure. They identified a growing desire amongst their customer base to interact via online channels. However, during incidents where the bank’s app or website login experienced downtime, calls to the contact center would skyrocket, potentially leaving customers hanging after phone lines closed for the day. The bank did not have a 24/7 call service, and they needed to control what was happening in the call center. They realized that a chatbot could be a viable alternative channel to manage the increasing demand for online chat as opposed to phone calls.
Íslandsbanki is one of the three major banks operating in Iceland, offering a full suite of services including corporate banking, asset management, and retail banking. With roots dating back to 1884, the bank has a legacy of pioneering future-looking initiatives that enhance the banking experience for its customers. The bank has been trying to make the banking experience more digital and less branch-heavy, while simultaneously improving its infrastructure. They identified a growing desire amongst their customer base to interact via online channels, which led to the introduction of an AI-powered virtual agent named Fróði.
Íslandsbanki decided to introduce an AI-powered virtual agent named Fróði developed on the boost.ai platform. The bank, along with leading Nordic consultancy Advania and boost.ai, ran a two-day proof-of-concept (POC) workshop in which they developed a basic version of the chatbot Fróði. This prototype could query a number of APIs, including a currency conversion tool, as well as answer some basic questions. The boost.ai platform comes pre-loaded with over 1,500 banking-specific topics that a chatbot can answer out-of-the-box, which helped to greatly reduce implementation timelines. The bank was able to layer its products and services knowledge over the top via boost.ai’s no-code conversation builder. The chatbot was trained to use the software in just two days. The bank is also planning to have the chatbot handle financial transactions and more advanced functionality via authentication.