Redesigning the Tutor
Operations System for Online
Education
In the rapidly expanding landscape of online education, the dual-teacher model has emerged as a powerful way to scale high-quality instruction while maintaining personalized support. In this model, a main teacher broadcasts to thousands of students, while Teaching Assistants (TAs) manage localized pods. As the Lead UX Designer at XDC, I was tasked with upgrading the 'Class Mentor' platform, the primary operational dashboard used by these essential TAs.
Background
Teaching Assistants are the unsung heroes of the virtual classroom ecosystem. They are responsible for monitoring student engagement, managing real-time attendance, and facilitating interactive exercises. The overall success of this educational model heavily relies on their ability to proactively identify and support struggling or distracted students during live broadcasts.
As our user base grew, the operational overhead placed on TAs became a critical business bottleneck. We needed to transform their highly fragmented, multi-screen workflow into a streamlined, all-in-one dashboard. The goal was to drastically reduce their cognitive load and improve response times, ultimately enhancing the learning experience for students and the platform's overall operational efficiency.
Problem
The legacy system was a fragmented, operational nightmare for our users. TAs were forced to manage their classrooms using over 9 different tools simultaneously. This chaotic tech stack included Excel spreadsheets for attendance, WeChat for parent communication, the proprietary 4S platform for data, and multiple disconnected video and chat dashboards. This setup created an unsustainable cognitive load during high-pressure live classes.
Problem to Solve
TAs were forced to manage their classrooms using over 9 different tools simultaneously — creating an unsustainable cognitive load during high-pressure live classes.
Research
The legacy system forced TAs to juggle more than nine disconnected tools during every live class. To ground the redesign in reality, we combined qualitative shadowing with quantitative validation across the TA population.
User Observation
Shadowed TAs during live broadcasts to document window-switching behavior and reaction times.
User Interviews
Conducted 1:1 sessions to understand the emotional and cognitive toll of the legacy system.
Quantitative Surveys
Surveyed the broader TA population to validate usage patterns of specific video and chat modules.
"I spend more time fighting with overlapping windows and spreadsheets than I do actually looking at my students"
Chenxue Liu — Senior Teaching Assistant
User research data
Tutor-side Module Usage
During live lectures, the most-used tutor-side modules are the interactive chat area (97.98%), opening the lead teacher's video (94.52%) and the teacher chat area (93.08%).
Early-childhood TAs rely more on student video monitoring (93.68%); upper-primary TAs on teacher chat (97.31%), student coverage stats (75.27%) and quiz-answer switching (72.04%); middle-school TAs on class ranking (86.36%), class PK stats (83.33%) and consecutive unanswered quizzes (83.33%).
Class-Monitoring Focus
The data TAs monitor most closely: attendance rate, total online and total present.
The student types they care most about: offline / dropped / idle (91.35%), quiz non-participation (79.25%) and frequent wrong answers (78.96%).
For positive behaviors, in-app public praise and messaging parents are the most common follow-ups; for negative behaviors, TAs prefer reaching out to parents — light in-app reminders are used least.
We also deployed quantitative surveys to validate our observational findings across a broader user base. The survey data revealed that core monitoring tasks were highly demanding but poorly supported by the UI. We found that 94.52% of TAs relied heavily on continuous student video monitoring, and 93.08% needed constant, uninterrupted access to the main teacher's video feed to maintain context.
Research Insights
Synthesizing our research, we categorized the core problems into three distinct buckets: functional, operational, and informational gaps. The primary insight was that TAs did not need more features added to their plate; they desperately needed a unified ecosystem. The constant context switching was actively detracting from their ability to mentor and guide students effectively.
- Consolidate functional windows to minimize context switching.
- Pre-class: Streamline attendance tracking.
- In-class: Monitor student engagement in real-time.
- In/Post-class: Facilitate parent feedback.
- Student Data: Attendance rate / Total online / Total present.
- Student Status: In-class learning engagement.
- Student Behavior: Positive reinforcement / Corrective prompts.
- Learning Content: Key concepts delivered by the instructor.
Design Objective
We established three core design goals to anchor the redesign process. First, we needed to create an 'all-in-one' monitoring mode that consolidated the disparate tools into a single source of truth. Second, we aimed to drastically improve operational efficiency by automating routine workflows like attendance tracking. Third, we wanted to enhance overall user satisfaction by reducing visual clutter and cognitive overload.
Unify
Consolidate fragmented tools into one workspace
Automate
Streamline repetitive operational workflows
Simplify
Reduce cognitive load with a clearer interface
Information Architecture & Layout Exploration
The key challenge was choosing the optimal dashboard layout. We evaluated multiple configurations (1:4, 1:1, 3:2, 1:3:1, 1:1:3, 1:2:2) based on information density, efficiency, and visual hierarchy.
Overly symmetrical layouts (e.g., 1:1) lacked a clear focal point, while margin-heavy designs increased mouse travel and slowed response to critical alerts.
1:1 (Symmetrical)
Low
Split Focus
2 / 5
1:4 (Side-heavy)
High
Margin heavy
3 / 5
1:3:1 (Center Focus)
Optimal
Centralized
5 / 5
Pre-Class: Streamlining Attendance Workflows
The pre-class phase—specifically the 15 minutes before a broadcast—was previously a chaotic scramble. TAs had to cross-reference an offline Excel roster, check the 4S platform for live logins, and then open WeChat to manually ping missing students. This manual reconciliation was error-prone and highly stressful right before a live session.
Legacy Pre-Class Workflow
Manual cross-referencing between Excel rosters, 4S platform logins, and WeChat messaging.
Redesigned Workflow
A single dashboard view with real-time attendance stats (e.g., 80/100) and one-click automated reminders.
We redesigned the workflow with a centralized dashboard and automated attendance system. Real-time attendance and interaction readiness are now visible at a glance, giving TAs a clear overview of the entire classroom.
Serves as the starting point of the visual scanning pattern, establishing a primary visual hierarchy.
Verify data metrics based on questionnaire results
To reduce manual follow-ups, we introduced one-click attendance reminders directly within the dashboard. TAs can instantly notify offline or inactive students without switching to WeChat, significantly improving pre-class efficiency and reducing operational overhead.
Live Session Monitoring
OA/4S
Wechat / 4S
Excel
Internal App
Status
Based on questionnaire data, the system consolidates the abnormal student states teachers care about most — Absent / Idle / Offline — into dynamic alert prompts.
- 15 mins pre-class: alerts for students who haven't shown up.
- In-class: alerts for offline / idle students.
Actions
One-click attendance / idle / offline reminders — send WeCom (Enterprise WeChat) notifications to parents.
In-Class: Student Engagement Monitoring
Student Analytics
Track learning metrics
Video Monitoring
Monitor student engagement
Student learning data
The learning analytics module provides teachers with real-time visibility into students' online status and classroom engagement.
After class, the system automatically generates learning reports and sends them to parents, ensuring clear and continuous tracking of student progress.
AI-Assisted Video Monitoring
Monitoring dozens of student video feeds is the TA's most demanding task. In the old system, small thumbnails and manual scrolling made it easy to miss distracted or struggling students.
The new 1:3:1 layout introduces an AI-powered video grid that automatically detects and highlights abnormal behaviors (e.g., absence, poor lighting, disengagement) and prioritizes these feeds for quick review — shifting the TA's task from passive scanning to active mentoring.
In-Class: Driving Engagement & Recommendations
Beyond passive monitoring, TAs are responsible for actively encouraging student participation. Our research showed that maintaining high interaction rates was crucial for learning outcomes, but TAs struggled to track who had participated and who was falling behind in real-time using the old, disjointed tools.
We designed a recommendation system that provides real-time intervention suggestions based on student performance. For example, when a student answers multiple questions correctly, the system prompts the TA to send a reward or encouragement message with one click.
Manual Monitoring
Track learning metrics
Automated Monitoring
Monitor student engagement
@ Mention in Chat
💬 Direct Message
💬 Direct Message (Audio)
Interaction Participation
Participation rate achieved during live interactive exercises using the new platform.
Interaction Accuracy
Accuracy rate maintained by students, monitored seamlessly by TAs.
Post-Class: Automated Reporting for Parents
After class, TAs need to summarize student performance and communicate with parents. Previously, this required manually compiling data from multiple platforms, resulting in a time-consuming workflow.
We designed a post-class reporting module that automatically aggregates attendance, participation, and behavioral data into a clear and standardized learning report.
"The automated reports save me at least an hour every day. I can finally give parents detailed feedback immediately after class finishes"
Jie Li — Lead Class Mentor
Interaction Design Guidelines
Impact & Results
The launch of the upgraded Class Mentor platform delivered transformative results for our educational operations. Most notably, we successfully consolidated over 6 disparate legacy tools into a single, unified interface. This massive reduction in context switching fundamentally improved the day-to-day experience and cognitive load of our Teaching Assistants. Achieved an NPS of 8.3 post-launch.
Teacher Satisfaction
Accuracy rate maintained by students, monitored seamlessly by TAs.
Improved Efficiency
Reduced task completion time by approximately 70%.
TA Productivity
Increased TA-to-student operational capacity.
Quantitatively, the new 1:3:1 layout and AI-assisted features led to a remarkable 3x improvement in video preview and monitoring efficiency. TAs can now identify and respond to student issues significantly faster than before. During live sessions, the platform successfully maintained an exceptional 97.98% overall interaction rate, proving that the new UI robustly supported high-volume engagement. Furthermore, specific engagement metrics within the new system showed strong performance, with an 89% interaction participation rate and an 85% interaction accuracy rate recorded in key sessions. The automated workflows dramatically reduced manual operational overhead, allowing TAs to focus their energy on genuine educational mentorship rather than administrative software wrangling.
Reflection & Next Steps
Leading this feature-focused redesign reinforced the immense value of deeply understanding specialized, high-pressure user workflows. By physically shadowing TAs and mapping their chaotic, multi-tool reality, we were able to design a solution that didn't just look visually modern, but actively solved their most pressing cognitive bottlenecks.
Moving forward, we aim to leverage more AI capabilities—such as using platform-collected data for prediction and guidance—to empower teachers to improve their efficiency and sense of accomplishment, while enhancing students' motivation to learn.