What Canada’s Public Sector Can Learn From Agentic AI in Practice

What could Canada gain from the agentic AI revolution? That was the question at the center of a recent national hackathon in Ottawa, where nearly 90 teams came together to explore how AI can be applied to real public sector challenges. The event brought together public sector leaders, industry experts, and builders to explore how AI can improve transparency, accountability, and outcomes across government systems.

With more than 600 participants registered, Bounteous team members traveled and participated as a team, exploring solutions, bringing in AI experience from our different industries; ultimately, building a procurement platform demonstrating what agentic AI can do when paired with meaningful data.

Making Government Data More Usable with AI

Canada already leads by example in making federal spending and grant data publicly available. But accessibility alone does not guarantee impact. These datasets are massive, complex, and often difficult to interpret without specialized tools or time-intensive analysis.

As part of the hackathon, the Bounteous team developed a working platform designed to make Canada’s public procurement data immediately actionable. Built on open federal and provincial datasets, the solution introduced an AI-powered interface that allows users to explore complex spending data through two primary lenses: identifying contract “amendment creep” and detecting vendor concentration trends. Instead of relying on static reports, users can ask questions, such as, “Which vendors have seen the largest growth in sole-source contracts?” and receive real-time, auditable results with full source transparency. Every insight is traceable back to the original dataset, ensuring both accuracy and trust while dramatically reducing the time required to uncover meaningful patterns.

With this platform, government leaders are enabled through AI. Procurement teams can proactively flag contracts that exceed defined growth thresholds before they become risks. Policy leaders can identify categories where vendor concentration is increasing and prioritizing intervention to maintain a healthy market. Executives can quickly answer high-level questions about spending patterns across departments without waiting for manual analysis. By enabling faster, more informed decisions at every level, the platform demonstrates how agentic AI can turn publicly available data into a continuous feedback loop, one that supports better governance, stronger accountability, and more effective use of public funds, ultimately turning publicly available data into a tool for continuous accountability.

These examples helped address the hackathon challenges, but are easily adapted to other industries. Many of these problems are universal, so bringing in examples from Bounteous clients and other industries can help reinforce best practices and uncover surprising areas of innovation.

The Role of Agentic AI

What made this hackathon distinct was the use of agentic AI or systems that go beyond simple automation to actively assist in reasoning, analysis, and decision-making.

Rather than querying static reports, AI agents are capable of exploring datasets, identifying trends, and generating hypotheses for further investigation. This enables analysts to focus less on data retrieval and more on interpretation and action. In addition, having agents “show their work” is imperative so that decision makers can vet the information themselves, evaluate the results, and ensure guardrails are in place. Observability in AI is incredibly important to enable these systems to operate more autonomously.

Designing for Real-World Constraints

A key takeaway from the event was that the most effective solutions were grounded in clear constraints. Teams worked with defined datasets, focused on specific use cases, and aligned their outputs to practical outcomes such as improved oversight or faster analysis.

This reflects a broader shift in how organizations are adopting AI. Early efforts often centered on experimentation. Today, the emphasis is on integration and embedding AI into workflows in a way that is measurable, repeatable, and aligned with organizational goals.

A Glimpse of What's Next

Canada is well-positioned to lead in applied AI, particularly within the public sector. With strong data infrastructure already in place and growing collaboration across provinces, industries, and disciplines, the foundation exists to move quickly from experimentation to implementation.

As the entire world shifts into execution mode, it’s clear that uncovering patterns and marking it easier to use data is only the first step. Organizations need internal processes, executive sponsorship, and the resources to act on those insights to improve processes and outcomes. By embedding AI into the core of how work gets done, rather than layering it on top, organizations can unlock efficiency, improve decision-making, and deliver better outcomes for citizens.

The work explored in Ottawa is a step in that direction and a reminder that the value of AI is not just in what it can do, but in how effectively it is applied.