SpainFinance and insurance2018 – ongoing

EVO Banco

Voice assistant that answers customer calls to the contact center

EVO Banco's voice assistant handles about 85% of contact center calls and cut average time on the line to two minutes

WorkingVendor case studyVendor ↗

Problem

EVO Banco, a Spanish digital bank, already had a virtual assistant in its app and on its website. It had no way to connect that assistant to its call center to answer telephone queries.

Approach

In 2018 the bank turned to Google Cloud and, with the partner Altostratus, built a voice platform. Call audio is transcribed by Speech-to-Text and sent to Dialogflow, which works out what the customer wants and either answers or passes the call to a human operator. Call data is processed to train machine learning models that improve the system over time.

Results

about 85%share of contact center calls handled by DialogflowCompany's own claim[1]

2 minutesaverage time on the line for an answer, down from 5 to 6 minutesCompany's own claim[1]

95%share of calls routed correctlyVendor case study[1]

Takeaway

The bank tracks routing accuracy and a second measure it calls conversational effectiveness, which stood at about 70%. It said the aim was to free staff from routine inquiries, not to cut calls to people.

Sources

EVO Banco: Developing a human-centric voice banking platform with Google Cloud
Google Cloud10 May 2022Vendor publication

Source summaries

[1]Google Cloud10 May 2022

Google Cloud's customer story describes how EVO Banco built a voice banking assistant for its contact center using Speech-to-Text and Dialogflow. It reports that about 85% of calls are handled by the assistant, average time on the line fell from 5 to 6 minutes to two minutes, and calls are routed correctly 95% of the time. The page carries no date; the Internet Archive first saved it on May 10, 2022.

Full details are in the original sources.

Added 4 Nov 2026 · Updated 4 Nov 2026