Operationalizing advanced analytics and AI at Scotiabank.

0

This is a five-part blog series drawn from an interview I recently had with Grace Lee, Director of Data & Analytics and Dr. Yannick Lallement, VP, AI & ML Solutions at Scotiabank.

Scotiabank is a Canadian multinational banking and financial services company headquartered in Toronto, Ontario. One of Canada’s Big Five banks, it is the third largest Canadian bank by deposits and market capitalization. With more than 90,000 employees worldwide and assets of approximately $1.3 trillion, Scotiabank has invested heavily in artificial intelligence, analytics and data and aligned an integrated function that is well supported by all sectors of activity. Although their journey has had a zigzag impact throughout, the organization is now well positioned to deliver continued value and impact to the business.

This five-part blog series answers these five questions:

BlogOne: How is the advanced analytics function structured and what have been some of the biggest operational challenges in your journey?

Blog 2: What does it take to set up an AI/ML Solutioning competence center?

Blog 3: How do some of the operational challenges like digital literacy impact your journey?

Blog Four: What are some of the operationalization lessons learned?

Blog five: What next for Scotiabank’s advanced analytics and artificial intelligence capability?

How do you operationalize integrated budget planning to ensure that business functions and your analytics/AI solutions remain integrated?

“First of all, we are at the table together. We mutually agree that we need to grow the business and serve our customers. From there, our data and analytics teams assess whether we are using AI for the solution or using something else. The challenge we face is not how to find value in AI or how we can build the most sophisticated models; it’s how we grow the business and find value for our customers, employees and shareholders. This is where our common goals become so important. We are not a hammer running around looking for a nail. AI is just another valuable tool in our toolkit to help the business achieve its goals. (Verbatim: Dr. Yannick Lallement).

How do you ensure communication and change management practices are applied to support your functional excellence?

We have processes in place to manage our capacity, which means we are careful in the projects we select. The key here is for the company to have a good understanding of what we can and cannot do. Data and analytics awareness initiatives are one of the ways we aim to increase understanding and knowledge across the Bank. For example, we host an in-house data and analytics week every year and have over 1,000 attendees from all corners of the organization learning about data, analytics and technology and what it can do. do and have done for the Bank. We also have regular communications and presentations to help the company continue to learn how using analytics can further improve their business. (Verbatim: Grace Lee).

What is one of your AI projects that you are most proud of?

During the pandemic, we developed a model to identify customers likely to experience financial hardship, or the Customer Vulnerability Index. Through this exercise, we were able to identify approximately 2 million customers who may need our assistance. The entire bank came together and created a response team to make calls and support our customers with proactive outreach. We informed our clients about special government assistance programs, as well as ways to redirect their portfolios so they are better positioned to weather the storm. Not only has this helped our clients where needed, but we have also been able to reduce default rates and reduce risk to the Bank. In addition, during this period, we have seen a five point increase in our clients’ Net Promoter Score (NPS), which reflects our clients’ appreciation for our care.

We are so proud of this initiative because we were able to use AI during the pandemic to better serve all of our customers. It was an unprecedented time where there was an enormous amount of uncertainty for millions of our customers at once. Serving them all in a personalized way can only be accomplished through AI, digital enablement and our people all working together. (Verbatim: Grace Lee).

How are AI ethics integrated into all your AI/ML programs?

We have created a function dedicated to Data Ethics, reporting directly to our Chief Data Officer. This team, in partnership with our privacy and risk functions, has developed an ethics assistant to support our goal of facilitating the right action, including for AI model releases. This is a comprehensive checklist to ensure our teams continue to focus on data ethics when building models or co-creating with the business. And it’s really like on a plane, you know, you have a take-off checklist, you go through everything to make sure your plane is ready to take off. In this case, the ethics assistant helps us to examine this.

Using the checklist, we may either need to come back as modellers and modify elements of the model that may create bias, or go to the company and discuss the implications. If there’s a bias, we always go back to the drawing board and take the company with us on the journey. This way, they begin to understand responsible data use in a much more practical way.

The checklist is also an important way to educate people on the things they need to consider to ensure our modeling activities are aligned with our core values. There is constant turnover in our field, with new people joining the team every day. Tools such as the Ethics Assistant help reinforce their training and development and give us confidence that we are using a consistent approach across the Bank. (Verbatim: Dr. Yannick Lallement).

Share.

Comments are closed.