BTEX 2025: The Microsoft Copilot Revolution Transforming Canadian Workplaces

Presanna Vingnanalingam kicked things off by sharing CAAT Pension Plan’s Copilot journey. “When I think about our journey, it started about a year and a half ago,” he explained. “We started reviewing a lot of our business processes and capabilities within our organization, and we started looking at specific bottlenecks that would constrain some of our objectives and key results as well as our growth.”

With Microsoft removing the 300-licence minimum, CAAT saw a timely opportunity. “We decided, hey, we’re going to hop right in, grab a few licences and try it out,” he recalled. Starting with 20 highly motivated users, the pilot soon grew to 60 users, spanning departments. The results were clear: “We saw on average about a 20 percent increase in terms of productivity across a number of different business areas.”

This success paved the way for a wider rollout. Within a year, CAAT added hundreds of licences and partnered with CDW to launch a deeper Copilot adoption initiative. “We started focusing specifically on different areas within the company like HR, Legal and our Growth teams,” said Vingnanalingam.

Sunny Wang offered a broader industry context, highlighting how the shift from individual to organizational productivity is accelerating AI adoption. “At Microsoft, we are now focused on team, department-level and organizational productivity.”

This shift, she explained, moves us from personal Copilots to organizational agents. “Imagine an HR agent that answers HR questions, takes intake from employees… this idea that a person manages a suite of agents. Humans with their own assistants, and those assistants have agents that can take tasks on for you.”

Wang noted the emergence of the frontier firm: “Every person gets their own Copilot. Now we’re moving into an era where every person has a team of AI agents helping them get work done.”

As Copilot adoption scales, organizations are also exploring more tailored solutions. Otto asked Wang to elaborate more on Copilot Studio:

“It’s a low-code, no-code way to build agents for your organization,” Wang explained. These agents can tap into both Microsoft and third-party systems. “You probably have Workday, ServiceNow, Salesforce… an agent should theoretically be smart enough to access all of that and then use generative AI to draft, write code, do all the things you're used to.”

She emphasized Microsoft’s tiered platform strategy. “Start with personal Copilot. Then, you move to M365 Copilot for organizational productivity. And for advanced needs, there’s Azure AI Foundry.”

At CAAT, moving beyond the basics meant turning to Copilot Studio. “Where we hit a little bit of a wall with just Copilot for M365 is the way that the data and the files were indexed,” said Vingnanalingam. “Sometimes you weren’t getting the responses you were looking for.”

To address this, his team started building their own agents for HR, Legal and investments. “We started using Azure AI Foundry to start creating specific agents for specific tasks,” he explained.

The goal: enabling end-to-end workflows that span multiple platforms. “We have a Salesforce investment, ServiceNow investment,” he said. “The question is, if we’re looking at an end-to-end business process that moves through a number of these systems, how do we enable an autonomous agent to talk to these systems and help drive that process?”

Proofs of concept are already in play. “We’re now POCing it across a few of our heavier processes – things like retirements, the onboarding of new employers onto the plan and the way we do annual statements,” he added.

Ahmed spoke about the variety of use cases emerging across industries. “There’s always that internal piece… things like IT service desk, HR, finance self-serve – very internal-facing,” he said.

On the external side, municipalities are using agents to assist the public. “If I'm a resident or tourist looking for information – what kind of dog licence do I need? What trails can I hike this weekend? That info is often hard to find on websites.”

The takeaway? “Once you identify the use cases, building the application is not that complicated at all,” said Ahmed.

When asked about the lessons he has learned, Vingnanalingam emphasized three best practices that helped CAAT build sustainable AI initiatives:

Start Small: “We started with a group of folks who are highly friendly. They’re curious, supportive and helped create momentum. We focused on time to value – how quickly individuals can realize benefits from Copilot.”

Data Security and Compliance: “Being in financial services, we have a lot of regulatory requirements. We're very careful but also very aggressive in our adoption approach.”

Governance and Culture: “We have an AI governance committee staffed by our senior executive team. We also have a gen AI policy in place to provide prescriptive guidance.”

Wang expanded the list with three additional insights from her work with customers across Canada:

Know Your Why: “Customers say, ‘This sounds cool.’ But is it about saving time? Attracting talent? Changing processes? If it’s not linked to business outcomes, it won’t last.”

Appoint an AI Project Manager: “Even if it’s a part-time role, having someone accountable helps vendors like Microsoft support you better. It gives structure and continuity.”

Invest in Skilling: “AI has exposed gaps in digital literacy. Not everyone is comfortable with Office apps. You have to lead users through the journey – not just hand them tools.”

She pointed out that even organizations that deployed Copilot widely saw uneven adoption. “Some picked it up fast. Others needed more support. We encourage customers to use the Copilot dashboard to track adoption and guide training.”

Vingnanalingam agreed. “We used Microsoft’s adoption kit for the first group. We now realize more structured training is needed to get people comfortable. We track metrics like usage by application to identify where we need to focus.”

Brian Matthews ended with a bonus question for the panellists: What excites you most about generative AI?

Vingnanalingam was the first to respond. “This is one of those technologies that has completely transformed the workplace – and my personal life too,” he said. “We planned a whole camping trip with my wife and son using AI.”

Ahmed offered a cultural observation. “I’m always surprised by how many IT departments are in the basement. But now, with AI, technology is mainstream. It's part of the top-level business conversation.”

Wang got personal. “Before Copilot, I was working evenings and weekends. Now I get my admin work done by 3 p.m. It’s changed my quality of life.”

She also shared an inspiring story. “There’s a 65-year-old trainer at a traditional investment firm in Toronto. She decided to learn Copilot and now the CEO calls her for help. She’s even thinking of consulting after retirement.”