What’s the best credit card for me? Can I reverse a bill payment? Getting the answers to these common customer questions is quick and seamless. That’s because Scotiabank’s data science team is innovating to enhance the customer experience.

Yannick Lallement, Scotiabank’s Director of Data Science, says the team is implementing machine learning — a branch of artificial intelligence where systems can learn from patterns in data — to make the content search experience faster and more accurate for frontline employees who assist with customer queries.

The team’s efforts led to Scotia Search launching in 2018 for online and branch customers, for retail and business service centres, and for contact centre employees in English and Spanish — and March 2021 recorded the Bank’s highest visitor traffic at 60,000 users and 500,000 searches monthly.

The team is now improving the Scotia Search experience by experimenting with an engine that understands language and is trained on trillions of words. The engine is called Generative Pre-trained Transformer 3 (GPT-3) and it’s created by artificial intelligence research lab OpenAI.

“We wanted to find out if the GPT-3 engine can help customers easily get answers to all their complex banking questions,” said Lallement. The team hopes to launch an interactive chat bot to instantly answer common questions using this engine.

To help with the experimentation and to encourage new ideas, the team hired Nicole Vavrukh, a third-year Simon Fraser University Computing Science student through the Bank’s Velocity program. Vavrukh worked on a use case for Scotia Search and the chat bot, and lately has been fine tuning search parameters to maximize the system’s accuracy.

Lallement thinks the chat bot will be a unique value proposition for Scotiabank customers because it will be able to search everywhere on Scotiabank.com — and not be limited to just specific content like other chat bots out there. “That chat bot currently doesn’t exist on the internet. It will be the first of its kind that answers financial banking questions based on knowledge published by a major bank,” he said.

“We’ve made great strides in how we’re using data to drive real value for the business,” said Grace Lee, Scotiabank’s Senior Vice President & Chief Analytics Officer. “Emerging technologies gives us a way forward to identify future use cases that can improve customer experience.”


Enjoyed this story? Stay tuned for more from our Innovation Spotlight series which showcases our leading-edge technology along the journey from code to customer. See our previous story:

A Big Mac and Scotiabank’s technology strategy are similar. Why? Think global. Think standardization.

May 12, 2021

In this first feature, we share how the Bank is powering local innovation and scaling it across borders