Have you ever sat through an AI presentation and thought:
“This looks neat, but what happened after it shipped?”
That reaction is common. Public AI conversations tend to reward polish, novelty, and clean success stories. POCs have never been easier to demo.
What’s harder to find is a credible place to talk about what didn’t age well, what failed quietly, and what had to be maintained long after the initial momentum faded.
For a healthy AI community to learn effectively, those experiences matter just as much.
In the Canadian enterprise context, these realities are familiar. Regulated environments, legacy infrastructure, and long-lived systems introduce constraints that rarely show up in demos. Agentic experiments are created by the dozen. Scaling them safely, reliably, and cost-effectively over time is a challenge entirely different.
That’s why the Toronto Machine Learning Summit speaker review process looks beyond what was launched and focuses on what happened after.
Our practitioner-led Steering Committee scans submissions for signals such as:
- Elegant architectures with no operating history
- Experiments with no clear ownership once live
- Results that haven’t faced sustained scale, cost pressure, reliability demands, or compliance realities
These criteria exist for a reason. Adoption lessons are only actionable if they reflect the real conditions teams operate under.
This year’s selection process places particular emphasis on:
- The metrics teams used to understand system behaviour over time
- The frameworks applied to manage scale, cost, and reliability
- The operational constraints that shaped decisions in practice
The goal is for each session to help the community understand how a system actually runs in production, not just how it was designed.
Over time, these case studies reveal clear patterns that practitioners can apply across common application areas and specific domains as they work to improve AI and agentic adoption.
Alongside production case studies, TMLS also features applied research from Canadian labs, presented in the context of how it informs real systems and decisions.
If you’re curious how this process works and whether your experience aligns with it, you can explore the Call for Speakers and review criteria in more detail. We’d love to have you submit!
→ Learn more about the Call for Speakers
Toronto Machine Learning Summit is practitioner-reviewed and focused on extracting durable, real-world lessons for AI practitioners, researchers, and executives. We’re looking forward 10th year of community insights and celebration of our biggest wins, failures, and lessons learned.
Thank you to our 2026 Steering Committee for the time and judgment they bring to this process.