From classroom evidence to an adaptive reading intervention

<aside> 📍

Role: Independent Product Designer, UX Researcher, and Product Strategist

Timeline: Discovery completed; MVP definition in progress. Target pilot: next academic semester, 2027.

Team: Independent project. Discovery evidence included 5 classroom teachers, 1 school principal, 1 vice-principal, and 1 pedagogical coordinator.

Tools & methods: Notion; contextual observation; diagnostic evidence triangulation; learning artifact analysis; Opportunity Solution Tree; MoSCoW-style scope decisions.

Environment: Public school context, with shared tablets available through scheduled use.

Product model: Student-facing mobile and tablet experience, paired with a teacher-facing progress dashboard.

TL;DR: I translated direct evidence from 24 students and school staff into a multidimensional reading-skill model, three prioritized opportunity areas, and an adaptive student and teacher MVP. Product validation has not started, so the current outcomes are discovery decisions and testable hypotheses, not claims of learning impact.

</aside>


Problem

Across grades 3 to 6, I observed students who could copy content from the board but could not independently decode it, read it back, or produce comparable writing without a model. A completed notebook could therefore signal participation while concealing limited decoding and autonomous writing.

As students progressed to higher grades, reading became an access barrier across the curriculum. The challenge was to identify where reading broke down for each student and define a digital intervention that could support practice without infantilizing, stigmatizing, or requiring students to read complex instructions before they could begin.


Discovery


Synthesis

The synthesis converted classroom evidence into a skill model, opportunity areas, and explicit evidence rules. The central shift was from assigning a general literacy level to identifying which skill breaks down, under which conditions, and with how much support.