Disco AI x PMC Product Spring
Turning Engagement Data into Timely Instructor Intervention
Role
Designer
Timeline
March 2026
2 Weeks
Team
4 Designers
Skills
Product Design
Product management
User Research
Rapid Prototyping
Overview
Disco is a modern, human-centric, AI-powered learning platform designed for organisations that deliver transformational learning experiences. With a growing customer base spanning bootcamps, training consultancies, professional associations, and corporate learning teams, Disco enables organizations to build, sell, and scale learning programs, run engaged learning communities, educate customers and partners, and upskill employees
The Problem: Instructors cannot easily detect which learners are disengaging until it is too late to intervene.
Instructors/program leaders can see that learners drop off, but they have limited tools to understand why. Additionally, instructors struggle to detect early warning signs when learners begin to disengage. Current learning platforms tell you what happened, but they never tell you who is about to quit.
This led to the question:
How might we surface clear, actionable signals about learner engagement so instructors can intervene before students drift away?
The At-Risk Learner Insights is an engagement diagnostics feature that surfaces and analyzes learner activity to identify students who may be disengaging or struggling. Unlike traditional analytics dashboards that require instructors to manually interpret engagement data, At-Risk Learner Insights translate engagement signals into clear risk indicators and explanations, enabling instructors to quickly identify learners who may need support.
The system works in 4 steps.
Risk Identification
It detects learners whose engagement signals indicate potential risk
At-Risk Learner List
Learners are then prioritized in a list from least activity to most.
Context
Insights then explain why those learners were flagged.
Actionable Suggestions
Helps instructors reach out with a suggested follow-up message, and offers solutions to improve the course content/structure
The Prototype
Below is a recorded demo of the prototype, lead by my teammate Carmen.
Key Features
Cohort Engagement/Key Insights
Instructors can view overall engagement trends along with key insights. These insights are also provided with context as to why certain trends are happening, as well as actionable steps an instructor can take to improve their outcomes.
At-Risk Learners
A list of learners that have poor engagement with the course, ranked from most likely to least likely to drop the course. Instructors can view their engagement signals, check-in surveys, a timeline of their activity, as well as the possible causes for that disengagement.
Instructors are also given a customized follow-up email template, allowing them to engage more efficiently with students.
Our Solution: At-Risk Learner Insights
Our Advantage: From Data to Early Intervention
What makes the At-Risk Learner Insights powerful is that it guides instructors through a clear progression
Early Detection
Instead of instructors manually scanning analytic dashboards, the platform will instead highlight learners whose engagement patterns signal a potential risk.
Diagnostic Insights
Diagnostic insights summarize the signals that triggered the alert, and the platform helps instructors quickly assess whether a learner may need clarification, encouragement, or additional support.
Instructor Intervention
With the diagnostic insights, the platform provides resources that allow instructors to quickly reach out and support learners before disengagement turns into drop-off.
This shifts the instructor workflow, making it more efficient by allowing them to focus less on interpreting data and more on supporting the learners who need help.
Upon interviewing online learning students and instructors, we found 3 key insights.
We interviewed both students taking online courses as well as 2 UBC Extended Learning Instructors to gain a better understanding of the problem space, and found key insights:
Interview Insights:
Lack of visibility into the learning process: Instructors primarily rely on assignment submissions or learner outreach to detect difficulties. Because most learning activity happens privately, instructors often cannot see when learners are confused or disengaging until the learner asks for help or submits an assignment.
Monitoring engagement becomes difficult as cohorts grow: As cohort sizes grow, instructors find it increasingly difficult to monitor each learner’s progress individually. Manually reviewing assignments, discussions, and participation across multiple learners makes it harder to consistently identify which learners may be struggling early.
Instructor feedback and support reinforce learner motivation: learners reported stronger engagement in courses where instructors actively monitored progress and provided feedback.
The at-risk learner's insight allows instructors to gain that missing visibility by providing important context as to why learner engagement is dropping. Instead of spending time interpreting engagement data, instructors can quickly detect, diagnose, and respond to learner disengagement — and understand whether the issue is individual or cohort-wide.
By transforming engagement data into clear signals and actionable insight, At-Risk Learner Insights helps instructors identify struggling learners earlier and intervene when it matters most.
My reflection
Not every problem needs a new fancy feature
Disco AI is already a very complete platform, offering instructors many ways to track learner engagement; such as as missed assignments, inactivity, quiz performance, and module completion.
A key thing I realized doing this case study was that the challenge isn’t collecting more information, it's helping instructors interpret the data already available. Because the most impactful and successful learning platforms aren’t the ones that show the most data — they’re the ones that help people act on it.