SchoolAI Insights

SchoolAI already generated valuable data around homework, student performance, subject proficiency, learning gaps and errors. But the existing Teacher Insights experience was primarily homework-focused. Information was difficult to process quickly, and teachers had to interpret the page before knowing what required attention.

Students and parents had no dedicated Insights experience. Students lacked a central place to understand progress and what to work on next; parents depended heavily on teachers and schools for updates. The opportunity was to turn existing data into an actionable learning experience, rather than simply another dashboard.

Rather than designing three independent dashboards, I approached Insights as a shared system. The same underlying student data could be interpreted differently depending on the user's role.

The data was highly interconnected. A teacher could begin with a class, move into a subject, identify a weak chapter, investigate a subtopic and eventually inspect individual incorrect answers. A student could follow a similar path through their own performance.

The hierarchy supports progressive movement from ‘What is happening?’ to ‘Why is it happening?’ without exposing every layer at once.

One of the biggest decisions was moving away from a single dashboard structure. Each experience instead began with the question that person was trying to answer. This shifted the design from data presentation to task-oriented exploration.

A subject teacher, class teacher and principal can look at the same school data but have different responsibilities. The interface adapted the scope of information while maintaining one underlying design language and interaction model.

Principal Insights · School summary with weekly activity, quick actions for appreciations and nudges, and alerts.

Principal · Homework completion by grade, subject and section.

Principal · Performance levels by grade, subject and section.

The redesign expanded the original homework-focused view into a broader understanding of class and student performance. Teachers could move from a high-level overview into the specific areas requiring attention without beginning in a dense data table.

Class Insights · Subject Teacher view of one subject across sections, from performance summary to chapter and subtopic levels.

Class Insights · Class Teacher summary with class and student highlights, homework completion and subject performance.

Class Insights · Class Teacher subject view with key highlights and chapter-to-subtopic performance.

Student Insights for teachers · Key highlights and a student list with performance levels and completion.

Student Insights was designed from scratch. Students were not monitoring other people; they needed to understand themselves. The experience focused on how they were doing, where they were struggling and what they could work on next.

The design connected reporting to potential next actions through a path from subject and chapter to a specific learning gap and recommendation.

Student Insights · Personal summary with actionable insights, completion, timeliness, level-ups and subject performance.

Subject view · Physics performance summary, key highlights and chapter performance with subtopics.

Gap Insights · Subtopic performance with the incorrectly answered questions behind each gap.

Parent Insights created a direct view into a child's learning journey. The goal was not to expose every metric available to teachers, but to surface overall progress, subject performance, areas requiring attention, learning gaps and change over time.

This let parents understand academic performance without needing to follow up individually with multiple teachers.

Not every piece of learning data needed to be visible at the same time. Each level was designed to answer a progressively deeper question while retaining access to the detail beneath it.

The hardest part was not creating individual screens; it was deciding which information belonged at each level. For every view, I evaluated the user's question, the information needed immediately, the context that must stay visible, what could be progressively revealed and which action understanding should enable.

Because the data was interconnected, changing one part of the structure could affect navigation, tables and drill-downs elsewhere. The work required repeated iteration across information architecture, metric prioritisation and role-based views.

The experience had to work with existing backend and API structures rather than assuming entirely new data models. The same student data also needed to serve students, parents, subject teachers, class teachers and school leadership at different levels of detail.

The work stayed within the existing SchoolAI design system while introducing patterns for dense, interconnected data. A decision in one view could affect navigation, tables and drill-downs elsewhere, so structuring and validating the experience came before visual execution.

The redesigned Insights system launched across Teacher and Student experiences, with Parent Insights extending visibility of student progress to families.

Principals requested additional dashboard capabilities following launch, indicating demand for broader school-level visibility.

This project reinforced that data-heavy products are rarely solved by simply improving the UI. The important work happened in sequence: understanding the user, structuring the data, defining the hierarchy, designing the flow, and only then designing the interface.

The same dataset can feel overwhelming, useful or actionable depending on how it is structured around intent. The goal was not to show less data, but to make the right data easier to understand, navigate and act on.

Good learning insights don't just tell users what happened. They help users understand what to do next.