Identifying students who need extra support is one of the most important responsibilities of an institution, and it's also one of the hardest things to do early. By the time a student is clearly struggling, the window for meaningful intervention has often already closed.
That's why we're introducing predictive insights in IgniteAI Ask Your Data, a feature of Intelligent Insights, now available with Canvas Next. This new AI-powered capability helps you anticipate where students are headed, not just see where they are today.
What's new
Ask Your Data can now generate predictions from your institution's historical data. Instead of asking only what's happening right now, you can ask forward-looking questions and get the answer back as a chart or table.
The initial release focuses on a single, high-impact use case.
Learner dropout risk
This prediction identifies learners who are likely to drop out. It draws on patterns like engagement, grades, assignment completion rates, and historical success factors, then flags learners whose trajectory looks similar to those who dropped out in the past — giving you a chance to step in before disengagement turns into departure.
More prediction types are on the way. We're starting with dropout risk because it's one of the highest-impact questions an institution can ask, and we'll expand based on what you tell us matters most.
Why it matters
Intelligent Insights already offers the Students in Need of Attention module (now also available in Canvas Plus), which surfaces key engagement and performance indicators so you can quickly identify at-risk students and take timely action — reaching out to the student, their teacher, or their observer.
Predictive insights complement that work, but from a different angle. Students in Need of Attention is institution-driven: it relies on a defined set of indicators packaged into a ready-to-use module that can further be customized and tailored to each institution. Predictive insights, however, leverages AI and predictive algorithms to surface forward-looking patterns based on your own questions and your historical data.
Used together, the two give users more ways to catch warning signs early, plan thoughtful interventions, and support student success across the institution.
Help us make predictions better
Predictions are only useful when they're accurate and actionable. As you explore this feature, your feedback directly shapes development of future use cases. When a prediction is helpful, pin it, download it, or give it a thumbs-up to tell us we're on the right track and help us improve the underlying model over time.
Predictive insights are available now in Ask Your Data, part of Intelligent Insights and Canvas Next. We can't wait to hear how you're putting them to work.
