SUCCESS STORY: EVO BANCO – GOOGLE CLOUD
EVO Assistant, the first voice assistant in Spanish in the banking sector
About EVO Bank
EVO Banco is a 100% digital financial institution that has been operating throughout Spain since 2012. It was created with the aim of reinventing finance to make life easier for people , betting on innovation, transparency and simplicity.
Currently part of Bankinter , today has more than 150 employees and has become one of the most important digital banking entities in the country.
Seeking the best care
Starting with the mission of Putting the customer at the center and providing them with the best service, the need arose to improve the way customers communicated with the bank. To achieve this, it was necessary to create an innovative tool that would adapt to the new technological reality and that in turn would not affect costs
The solution, a conversational AI beyond a Chatbot
EVO Banco wanted to become the Voice Banking and offer much more than a traditional chatbot. For this, they needed a Voice Banking service that would allow users to communicate with the bank easily and receive fluid and natural attention, as if it was a real agent who was on the other side.
In order to achieve this, They did a comprehensive assessment of all the latest technologies available on the market and Google Cloud turned out to be the right provider to do this with their Premier Partner, Altostratus Cloud Consulting.
The result, EVO Assistant
The project focused on developing a telephone interface to connect the new EVO virtual assistant to its traditional form of contact, which was the telephone call. In this way, any call received would be quickly answered by the new assistant, also ensuring maximum operational efficiency for EVO Banco.
That’s how EVO Assistant</ was born b>, a telephone interface in the cloud created with Open Source technology and based on Google Kubernetes Engine, which allows automatic scalability, as well as the ability to analyze user behavior in real time.
With EVO Assistant, bank customers can know their movements, execute their operations and solve their doubts in a personalized and specific way
The system creates a connection with the old technology of the call center through the assistant, giving a virtual and automatic response to process calls, 24 hours a day.
With EVO Assistant, the bank’s customers can know their movements, execute their operations and solve their doubts in a personalized and concrete way, while the system controls the conversation and the context of it, since that is able to understand natural language, and at the same time, extract knowledge from large amounts of unstructured data and learn from the interactions. For example, if a customer is dissatisfied, the system will detect it by his tone of voice and automatically refer him to a real agent.
The service thus became the first voice bank assistant in Spanish with artificial intelligence and a multiplatform system (web, mobile , Google speakers, phone…).
AI, ML and Google Cloud Kubernetes, the big players
In the technological field, the challenge consisted of creating a tool that would connect with the traditional technology of a contact center. To do this, using the open source technology of Kubernetes and PBX (Private Branch Exchange) was the best alternative, since it allowed them to make this connection and also provide an automatic response to demand, favoring scalability.
Specifically, in Google Kubernetes Engine a group of PBXs (old switchboards) process customer calls and from here, the system allows you to create a connection with the old call center technology through the assistant giving a virtual response and to process calls.
EVO Assistant is possible thanks to an orchestrated system that connects Google Speech to Text technology, which allows audio messages to be transcribed online in a call, and Dialogue Flow, the natural conversational AI with next-generation virtual agents from Google, which provides very varied responses to customers. This technology makes it possible, for example, if a bank user asks about the condition of their accounts at different times, the answer will also be different. The system can, in this way, control the conversation and its context in real time.
Pedro Tomé, Head of Disruptive Innovation, Big Data & Advanced Analytics at EVO Banco ensures that the virtual assistant was the best solution to meet the objectives of the financial institution.
“Evo Assistant reduced time operator response time from 6 to 2 minutes, increased the success rate of the conversational system and managed to contain more than 70% of calls without transferring them to agents.”
Pedro Tomé
The impact on the business
On the one hand, it has made it possible to improve customer identification and security, providing high scalability and flexibility through the use of Kubernetes from Google Cloud. operator response from 6 to 2 minutes and a 95% success rate was achieved in the conversational system created.
As an added benefit, the use of EVO Assistant represents a significant reduction in costs, representing only 3% of the real cost of a traditional call center.
The collaboration of the partner, key in the whole process
Altostratus’s role was to recommend and integrate almost entirely the Google Cloud product portfolio into this entire project analytics
The Altostratus team proposed to the client a solution that could connect to their architecture, taking advantage of AI to overcome the limitations of traditional solutions, and that would allow automatic scaling of the tool and the ability to analyze behavior of the user in real time
