Data Foundations to Better Serve Canadians

with Ima Okonny

Insights from Ima Okonny, Chief Data Officer at Employment and Social Development Canada.

Okonny began by explaining her role, “I work for Employment and Social Development Canada, and I’m Assistant Deputy Minister and Chief Data Officer here. I do my best working across the organization to leverage data to really enhance services to all citizens, regardless of where they live in this country, from coast to coast, and really driving value out of data responsibly, ethically, and leveraging whatever technology we can to really enhance service delivery across this great country.”

Okonny shared some of the complexities of the role and the breadth of people she works with and supports. “We do have a big population. The latest estimate said we sat at about 40 million. So what that means is we need to consider the different segments of the population, urban, rural, and make sure that we’re really meeting people where they’re at.”

“To be able to do that, you really need to leverage data to understand the population, to understand the diversity in the population, to understand the geographical locations of where people are at, and really to understand the preferences people have in terms of how they want services delivered to them.”

Building a passion for data

Okonny explained when data had started becoming a passion for her, “I started about 24 years ago in public service. And I was very fortunate to work in an HR department and I was tasked with working on compensation. And as we did this, I worked with a talented group that really cared, to make sure that we were making these calculations correctly, because if these calculations were wrong, what it would mean is people would not be able to pay their bills, people would struggle in terms of the benefits they deserved, and people would not be able to get their needs met for their families.”

“So what that taught me was I needed to be very intentional. I needed to really put the people who would be impacted by the actions I was taking into consideration and understand the impacts, the potential harmful impacts, the positive impacts, and the full spectrum implications of every decision that we made, you know, as we leverage data to make those calculations.”

Using new tools to help meet people’s needs

Okonny shared an example of the team have used data to enhance service delivery, “One of the things that really stick in my mind is a project we did where we leveraged natural language processing to find vulnerable and at-risk Canadians, and give them some of the benefits they deserve, so that we reduced a lot of the administrative burden on them. Instead of them having to come to us and apply, and fill out all these forms, we were able to leverage data, leverage natural language processing, and meet their needs in a very timely fashion.”

“And the beauty of it is that it followed all the protocols in terms of the privacy considerations, legislative considerations. We worked very closely with our program colleagues to make sure that everything we were doing was properly contextualized, and also very intentional in terms of looking at some of the potential risks. We mitigated risk to make sure that people would not be harmfully impacted by the tools we leverage to get this project on. And it also opened our eyes to the potential of scaling some of this work in a way that would really yield concrete benefits to Canadians across the country.”

Okonny expanded on the lessons she and the team learned from the project “One of the things we learned was the importance of really contextualizing the problem we were trying to solve, and then also contextualizing the data and bringing in legal, bringing in privacy from the conceptual phase… And that makes us go faster. The more we can ensure that we’ve built some of those foundational structures and the protocols in place from the start, we can move faster, and we can scale.”

Okonny explained that integrated approach also proved valuable in building data literacy across other teams within the department, “As we included people in the conceptual phase of the design and walking through the problem, what we found is it led to people’s data literacy increasing because they saw the challenges we’re trying to work through; things like data cleansing, data integration, understanding the business context of the data, and it exposed people to the what-it-takes story to get this right. So that was one of the unintended benefits of the whole initiative.”

Building data literacy across the organization

Okonny explained how building data literacy across the organization has become a priority for her, “Within my organization, we have built an enterprise data literacy program. We did a benchmarking study where we surveyed people across the organization to understand the different data literacy levels. It was very eye-opening. And based on that, we’ve been able to really focus on key areas, things like data management practices, and to really get people to understand not just the technical pieces around data, but why it’s important in the context of service delivery.”

“I’ve found there’s a lot of interest for people to understand what it means to them. So one of the challenges I’ve seen when people talk about data literacy, they come at it from a very technical perspective. And if you’re working in operations, or you’re working in policy, what you really want to know is how does this apply to me and my job. I think when you contextualize data literacy, for the end user, the person who’s actually going to benefit from data literacy, it makes it easier for them to adopt and really learn. But the challenge is making sure that you contextualize it.”

Data foundations for responsible use of AI

Okonny described how the team are laying the foundations for the responsible use of AI, “One of the practical things that we’re doing, is building a solid data foundation platform. And what that means is we’re embedding a lot of the governance procedures into our data foundation, so things like data catalogs, data link, a way to really audit some of this data, who is accessing the data. Do we have the right legislative pieces around this data? Do we have the right privacy protocols around this data? So we’re embedding it into our infrastructure, and then as we’re leveraging this data in AI, we’re showing that some of those governance pieces flow through what we’re doing.”

“For example, we have analytical labs in our infrastructure that are monitored. We have audit procedures around them. And we’re also making sure that we’re building fairness models and fairness frameworks that really enable us to interrogate those models based on our population.”

“A practical example is we have an indigenous population in Canada that’s quite diverse. And as we leverage the data, to improve service and enhance service delivery, we want to make sure that our models are representative of the population. So we’ve actually built a model that can assess fairness and tell us if our tool is going to have potentially harmful impacts on the population.”

“We’ve also looked at things like does the business context reflect the reality of the data we’re seeing. And does the interpretation of the policy intent, is it reflected in some of the results we’re seeing? At the end of the day, we want to make sure that the outcomes are aligned with really helping Canadians and getting them the services and the benefits that they deserve.”

Advice for other public sector leaders

Okonny shared advice for other leaders in the public sector who are considering using AI and other new technologies. “It’s very important to center this around the people we serve. It is very important to understand the problem we’re trying to solve. It is very important to be inclusive. To include people who don’t typically sit in a chief data officer function, to include people in the policy side, in the legal side, in the privacy side, in the program side, in the operational side because that is how you understand the context, to make sure that whatever you’re doing is representative of the clients you serve, because there is no way that one person understands all these considerations.”

“So it is important to bring people in to have the conversations, to understand the context, and to take advantage of the art of the possible, because there are opportunities for us to really shift how we do things and really get people to better outcomes. There’s so much opportunity. But for us to understand those opportunities, we need to understand the context we’re working in, and we need to find those leverage points that we can really push to have impact on people.”

Opportunities to use AI to solve ‘wicked problems’

Okonny concluded by reflecting on the opportunities she sees looking around the world. “I’ve been very inspired by the countries that have really leveraged AI for good. Countries that have really taken an approach of saying things like, we have populations that are underserved. So how do we leverage AI, and how do we leverage some of these technologies to solve some of these systemic issues that we’ve dealt with for decades?

“There are some countries, some organizations, especially in Europe, that have taken a look at this and have tackled some of this. Those are some of the countries that I’m looking at and those are the organizations that I’m looking at in terms of saying, maybe we can turn this around and let us leverage technology to solve some of these wicked problems that we’ve been dealing with for many decades, and finally solve them and get people to better outcomes.”

Listen to this episode on any of these podcast platforms:

Spotify

Apple Podcasts

Deezer

Libsyn