Matooto 🦜 Taglines: * Matuto nang may pag-unawa. * Here in Matooto, I can learn, and you can too. Matooto guides you through, so you can understand it too.
Inspiration
The name "Matooto" is a portmanteau of the Filipino word "Matuto" (to learn) and the Cockatoo, a bird famous for its ability to mimic, learn, and understand language.
We looked at two distinct, interconnected crises in the Philippine education system: the struggle with language and reading comprehension among students, and the severe administrative burnout experienced by teachers in under-resourced classrooms. We realized that if a teacher spends hours writing lesson plans, grading quizzes, and generating reports, they have less time to actually teach.
We built Matooto with a clear standpoint: give students the flexibility to assess and improve their reading comprehension anytime, anywhere, in their mother tongue—while simultaneously automating the heavy administrative lifting for educators.
What it does
Matooto is a dual-interface mobile ecosystem designed specifically for the Philippine classroom.
For Students:
Bilingual Reading Assessments: Students take a first attempt in English, selecting unfamiliar words to build their personal "Bag of Words." They then take a second attempt in their Mother Tongue (Tagalog, Cebuano, or Ilocano) to build true foundational comprehension.
The "Kamustahan" Feature: After assessments, an overlay captures the student's emotional sentiment. The system correlates this with their scores to give holistic context to their learning journey.
Gamified Learning: A customizable "Streak Pet" levels up based on reading time and assessment scores, keeping students engaged.
Pre-Made Library: Instant access to general education, science, math, and language resources.
For Teachers:
AI-Powered Assessment Generation: Teachers can define the length and grade level of a reading exercise, and Matooto generates a custom story, translates it, and creates multiple-choice questions automatically.
Smart Dashboard & Metrics: Teachers see real-time metrics for their classes, instantly flagging students who possess low reading comprehension or negative sentiments.
Automated Admin Work: The app automatically generates and exports crucial documents, including parent consultation letters, student comprehension reports, and class records.
How we built it
We architected Matooto as a robust monorepo to handle both the student and teacher experiences seamlessly.
Frontend: We built a cross-platform mobile application using React Native (Expo) and TypeScript, styled with NativeWind/Tailwind for a clean, responsive UI that works perfectly on the Android devices predominantly used in public schools. We utilized Expo Router for secure, role-based navigation.
Backend & Database: We leveraged Firebase for rapid development. Firestore serves as our NoSQL database, structured with heavy data denormalization so that complex dashboards load in a single, cheap query. Firebase Auth handles secure student and teacher logins.
AI Integration: The core automation is powered by the Gemini API running inside Cloud Functions. Gemini handles the heavy lifting of story generation, dynamic questionnaire creation, and accurate Mother Tongue translation.
Challenges we ran into
Building a dual-sided marketplace app in a hackathon setting is incredibly difficult.
Role-Based Routing: Ensuring that students and teachers were securely routed to entirely different dashboards immediately upon logging in via Google OAuth required tricky asynchronous state management.
NoSQL Schema Design: Designing a Firestore schema that could connect a student's dual-language assessment attempt with their Kamustahan sentiment—without requiring 30 separate database reads—forced us to rethink traditional SQL data structuring.
Mobile UI Quirks: Navigating the quirks of mobile development, such as overlapping safe-area views, invisible floating action buttons, and cross-platform Flexbox rendering, tested our frontend patience.
Accomplishments that we're proud of
Successfully implementing a true bilingual assessment flow that seamlessly transitions a student from English to their regional Mother Tongue.
Getting the Gemini API to reliably output formatted educational content (stories and quizzes) based on strict teacher parameters.
Creating an interface that doesn't just digitize schoolwork, but actively cares for the student through the Kamustahan sentiment tracker.
Building a working, deployable mobile application that can run on affordable Android devices right now.
What we learned
The Power of Duplication: We learned that in NoSQL databases like Firestore, duplicating small amounts of data (like user names inside a class array) is a best practice for incredibly fast UI rendering.
AI as an Assistant, Not a Replacement: We learned how to constrain LLMs to act as administrative assistants for teachers, giving educators the final say to edit and confirm generated stories before assigning them.
Ruthless Prioritization: We learned how to cut features that look flashy (like iOS specific builds) to focus entirely on the platforms and tools our target demographic actually uses.
What's next for Matooto
Cloud Vision Integration: Allowing teachers to simply snap a photo of a physical textbook page and having Matooto automatically extract the text, translate it to a Mother Tongue, and generate a quiz.
Expanded Languages: Adding support for more regional dialects across the Philippines.
Advanced Gamification: Expanding the "Streak Pet" into a fully interactive digital companion that guides the student through their daily reading habits.
Built With
- firestore
- matooto-is-a-cross-platform-react-native-(expo)-app-written-in-typescript-and-styled-with-nativewind.-it-utilizes-a-serverless-firebase-backend-(auth
Log in or sign up for Devpost to join the conversation.