Banco Falabella, the financial arm of Chilean retail giant Falabella, launched a new customer service chatbot last month at the Asobancaria financial conference in Colombia.
Called “Sofia,” the technology uses an artificial intelligence that the company says will be able to resolve issues in under a minute through Facebook chat that, in the past, have taken up to a half-hour to resolve through other channels offered by Banco Falabella, which has branches in Chile, Colombia, and Peru.
“We are launching a great innovation, one that we think will definitely change the user’s experience — and how they really live and interact with the banking industry,” said Harold Martínez, innovation and development manager at Banco Falabella, at Asobancaria 2017 in Cartagena.
Through machine learning, Sofia will also, in time, understand more complex issues and improve both its accuracy and response time to facilitate payments, provide balance statements, and give credit limit updates.
Specifically, Sofia will be designed to quickly answer questions about, among other things, payment deadline dates, the location of the nearest branch, and what Banco Falabella financial products are currently eligible for special offers at the company’s stores, including Falabella, Homecenter, and Makro.
The company has previously offered customer service through Facebook. But Martínez said that this channel relied on human representatives and could take up to 15 minutes from the time the customer asked a question to when the problem was resolved.
Even worse, this line of communication was not available for support 24/7. Robots, on the other hand, never sleep. “Sofía carries out these processes in less than a minute, which accounts for an efficient customer service strategy with strong results,” said Martínez.
As with most companies investing in customer service automation, Banco Falabella sees its artificial intelligence gains as a way to free up representatives to focus on higher-level tasks. “This will allow us to focus our human resources on following up on more complex questions or procedures,” said Martínez.
Photo credit: EEIM