TD Bank Group (TD) announced two pilots to support more streamlined customer experiences in its contact centres and accelerate the coding workflow for its engineers. When contact centre colleagues require support to answer a customer inquiry, they seek the advice of supervisory support teams. This process can lead to longer hold times. TD developed a generative AI virtual assistant that, once implemented, is designed to help contact centre colleagues retrieve answers to these day-to-day banking inquiries in seconds, resolving them in a shorter amount of time.

Layer 6 helped develop the generative AI virtual assistant and train it alongside colleagues with a deep understanding of TD policies and procedures. The generative AI virtual assistant provides summarized responses using conversational language to contact centre colleagues, including links to the TD policies and procedures used to source its answers. Colleagues can reference them and complete the necessary next steps to answer customer inquiries faster.

The introduction of large language models (LLMs) in contact centres has already proven to be impactful, according to a 2023 Stanford and MIT study. The study found that one Fortune 500 software firm's contact centre saw a 14% increase, on average, in the number of customer issues resolved per hour. The technology had the greatest impact among novice workers, who were able to solve 35% more customer issues per hour.

The Bank is testing the generative AI virtual assistant with front-line contact centre colleagues this year. To augment the work of its engineers, TD launched a pilot for GitHub Copilot, a generative AI programming assistant developed by Microsoft. GitHub Copilot is designed to help streamline simple tasks and accelerate code development.

Users can prompt Copilot in natural language to get responses to coding questions, explanations of complex coding topics, suggestions for security recommendations, and more. At TD, GitHub Copilot is assisting engineers by analyzing the code they're writing and providing real-time suggestions for completing and testing it. With additional time, TD engineers can focus on tasks that require more expertise and provide greater value to the Bank.

The platform also enables a more efficient developer experience by helping engineers develop maintainable code. Between September and December 2023, TD launched a pilot among a select group of engineers, who tested GitHub Copilot to help develop unit testing code. In a survey of colleagues who participated in the test, TD found: 50% of engineers surveyed said GitHub Copilot could save them up to 20 hours over a two-week sprint.

93% of data engineers surveyed said they were equally or more productive with the help of GitHub Copilot. 75% of engineers surveyed said it was easy to integrate GitHub Copilot into their everyday work. Both the contact centre and the engineering pilots were developed through TD Invent, the Bank's enterprise approach to innovation.

Within this approach, TD explores how emerging technologies like generative AI and spatial computing can be used to address real challenges before reviewing their potential to be scaled. The Bank has a robust innovation pipeline, spanning across multiple lines of business.