Inspiration While participating in the 3MTT and ALX programs, I became aware of how powerful AI tools can be in transforming education. But many children in Nigeria’s rural communities still lack access to quality education, stable internet, and modern tools. I wanted to build something that works without the internet, uses avatars to teach, and makes learning engaging, fun, and accessible to every child — no matter their location or background. What it does AvatarLearn is an offline-first mobile learning app powered by AI avatars that speak, teach, and guide children through lessons in math, English, science, and local culture. It works completely offline after the first setup, using voice, visuals, and animations to engage learners in a personalized way. How we built it Built the app prototype using Flutter for cross-platform support Used AI tools to generate character avatars and educational content Integrated Text-to-Speech (TTS) for spoken lessons Designed with a child-friendly interface Used local storage to make the app usable offline Challenges we ran into Making sure all features work fully offline Generating lightweight but expressive avatar files for low-end devices Localizing content for rural Nigerian learners Limited access to physical test devices and connectivity during development Accomplishments that we're proud of Successfully created an offline-learning experience for children Designed a friendly and interactive avatar learning interface Built a working demo with minimal tools and maximum impact Developed a vision for scaling access to education in underserved communities What we learned AI can help bridge the education gap — but only with smart, inclusive design Offline-first development requires careful planning and creative problem-solving Avatars and voice can greatly enhance how children engage with lessons Social impact solutions are powerful when they’re simple and accessible What's next for AvatarLearn: Offline AI for Kids Expanding the lesson library to include more subjects and local languages Adding interactive quizzes and gamified learning Partnering with schools and NGOs to distribute to rural areas Seeking funding or mentorship to scale the solution across Africa Creating a version for tablets and shared classroom use

Built With

  • dart
  • figma
  • flutter
  • generative-ai-tools-(e.g.
  • haiper)
  • local-storage-(sqlite
  • shared-preferences)
  • text-to-speech-(tts)-apis
  • vidnoz
Share this project:

Updates

posted an update

We’re excited to share that the first version of AvatarLearn: Offline AI for Kids is now complete and submitted for the Africa Deep Tech Challenge 2025! What's New: Flutter-based offline AI learning app for kids AI avatars using local NLP for interactive education SQLite storage for content access without internet Kid-friendly UI for rural and low-resource environments Watch our demo Explore the code We designed this to bring quality education to children in areas with limited connectivity. Stay tuned — more improvements and features coming soon! Feedback and collaboration are welcome!

AI4Good #OfflineLearning #FlutterDev #SkynetCafe #AvatarLearn #AfricaDeepTechChallenge

Log in or sign up for Devpost to join the conversation.