Aprova

Turning Educational Inequality Into Opportunity: A Mobile-First Learning Platform

PROJECT OVERVIEW

AI-driven platform making ENEM prep accessible for low-income Brazilian students. This capstone project for my Digital Design course demonstrates end-to-end product thinking—from research and competitive analysis to prototyping and usability testing.

CONTEXT

Brazil's ENEM exam determines university access. 60% of the 4-5 million annual test-takers are from low-income families without resources for adequate preparation. Additionally, 85% of students never receive adequate feedback on essay writing—a section that can invalidate an entire exam.

MY ROLE

UX/Product Designer

CHALLENGE

Existing EdTech platforms either have poor UX with advanced AI, charge prohibitive prices, or are too generic for ENEM prep.

SOLUTION

Designed APROVA—a mobile-first platform combining adaptive AI, instant essay feedback, and social learning to make quality ENEM preparation accessible for low-income students.

RESEARCH INSIGHTS

METHODOLOGY

Desk research analyzing government data (IBGE, INEP, MEC), academic publications on educational inequalities, and EdTech market reports (2020-2025).

KEY FINDINGS

Connectivity paradox

89% own smartphones, but only 41% have consistent broadband

Feedback gap

Public schools average 1 teacher per 28 students, limiting guidance.

Low self-efficacy

62% show low academic self-efficacy despite high motivation.

Inefficient study

73% use passive study methods vs. 23% using active learning

Mobile-first reality

78% study exclusively via smartphone (small screens: 4-5 inches)

MARKED GAP

Brazilian AI platforms are too complex. International platforms have good UX but are generic. No platform combines: ENEM-specific + intuitive UX + genuine freemium + mobile-first.

KEY DESIGN DECISIONS

MOBILE-FIRST ARCHITECTURE

Every screen was designed for smartphones first. This respects the reality that 78% of users study exclusively on mobile. Reduces cognitive load and removes barriers that intimidate users with low digital literacy.

Splash screen

INCLUSIVE ONBOARDING

Collects demographic and educational data to build user profiles and enable adaptive learning. Personalizes content recommendations and difficulty levels through data-driven iterations.

Onboarding screen 1 - personal info

PROGRESSIVE DISCLOSURE

Dashboard shows overview cards (progress, weaknesses). Detailed breakdowns accessible on demand. Balances at-a-glance usefulness with deep-dive capability on small screens.

Home dashboard

PLATFORM STRUCTURE: STUDY, PRACTICE, WRITING

Divides learning into three distinct sections:

Study - video lessons and documents (embedded content);
Practice - adaptive question bank with difficulty scaling;
Writing - essay feedback with AI analysis of ENEM competencies.

"Study" screen

THREE-LAYER AI

Predictive personalization — Adapts difficulty following Item Response Theory
Immediate feedback — Essay corrections with ENEM competency analysis
Contextual assistant — Chat support for navigation and study strategies

Feedback after an incorrect answer.

GAMIFICATION FOR SELF-EFFICACY

Points, levels, and badges compensate for missing feedback in traditional schooling. Visible progress systems help students with low self-efficacy.

Profile screen, with performance details, as well as goals, achievements, and progress history.

VALIDATION

TESTING

Usability testing with 2 target users.

POSITIVE FINDINGS

Users quickly understood platform purpose
Visual design motivated engagement
Study/Practice/Essay separation was intuitive
Achievement system drove engagement

CRITICAL FINDING

Font sizes were too small on 4-5 inch screens. Users struggled reading in low-light environments.

ITERATION PRIORITIES

Increase base font sizes
Implement responsive scaling
Test contrast under varied lighting
Expand testing to 4-6 users

RESULTS

DESIGN IMPACT

Validated the market gap: ENEM-specific + good UX + freemium + mobile-first
Designed inclusive onboarding with adaptive personalization
Created AI architecture that democratizes instant feedback
Documented insights for next iteration

KEY INSIGHT

SOLVING FOR ACCESS REQUIRES DESIGN THAT MEETS USERS WHERE THEY ARE

The platform succeeds through human-centered design that acknowledges real barriers (limited broadband, no tutorial feedback, low self-efficacy) and creates genuine accessibility: mobile-first, freemium without artificial limits, and immediate support systems.