NatWest group is offering customers services through speech recognition technology that will “better understand” and serve them.
Using Nvidia’s artificial intelligence (AI) technology, the bank’s voice-based customer service tool will be able to understand different accents and pronunciations, as well as everyday language concerning money.
The system, known as the Spending Insights Tool, is being tested with 75% of the NatWest Group’s Royal Bank of Scotland customers.
It has been developed by combining Nvidia’s technology with NatWest’s AI research and customer data. “Together, researchers and engineers from both organisations set out to address a challenge facing the next generation of conversational banking services: improving how speech AI understands the wide range of accents, words, phrases and financial language used across the UK,” said the bank.
Spending Insights allows customers to talk naturally about finances using voice and text. They can ask questions about their spending and receive personalised insights. “Customers can interrupt, ask follow-up questions and provide additional context, with the experience combining conversational responses with interactive charts and visualisations,” said NatWest.
Maja Pantic, chief AI research officer at NatWest Group, said customers can interact with the bank “in a way that feels natural to them”.
“If they choose to speak to AI, the technology needs to understand the different accents, words and everyday language used across the UK,” he added.
According to Aser Blanco, global head of banking at Nvidia, the real-time nature of speech recognition makes it challenging. “Working with NatWest has allowed us to apply Nvidia Nemotron open-source models to those real-world banking challenges and identify where the technology can be improved,” he said. “The results show how adapting speech applications to real-world language can make voice experiences more accurate and natural.”
Nvidia Nemotron automatic speech recognition converts spoken words into text, which is then interpreted by NatWest’s proprietary small language models.
It was initially tested with 90 NatWest members of staff from different parts of the UK with a variety of accents.
They were asked to not just read scripted questions aloud, but to ask them in their own words, said NatWest, “using the phrases, pronunciations and more informal or colloquial language they would naturally use when talking about their finances”.
NatWest then expanded the work using anonymised contact-centre transcripts, which it said enabled the model to “learn from a much broader range of real-world banking language”.
Voice banking is becoming an increased focus for banks, and they are applying AI to the challenge.
During a recent presentation reported by Computer Weekly, Bahadir Yilmaz, chief analytics officer at ING, said: “Conversational banking, being able to chat with your bank, is going to be a fundamental experience that people will be looking for in the next five years.
“What [ING has] right now is the first baby steps for the conversational AI chatbot, like a little personalised information, problem solving and being able to take an action,” he said. “But we are not that far away from heavy core capabilities, like sending and receiving money, applying for products and getting advice within the boundaries of regulations.”
Last month, digital challenger bank Zopa removed more manual tasks by offering customers the ability to bank through natural speech.
Its customers can now use voice commands to instruct an AI-powered personal digital banking assistant to manage everyday banking tasks, such as making transactions or setting savings goals.
Ask Zopa, as the feature is known, understands and acts on instructions given by customers using conversational language.
