Maestro — Intelligent Tutoring System
Adaptive AI tutoring platform built to personalize student learning paths in real-time and deliver study Q&A using Flutter Web, Riverpod, Reinforcement Learning, and LLM APIs.
| TechnologyTeknologi | Role in SystemPeran dalam Sistem |
|---|---|
| Flutter Web | Cross-Platform UI |
| Riverpod | State Management |
| Reinforcement Learning | RL Policy Engine |
| LLM APIs | LLM Inference Engine |
Fixed online course sequencing and lack of real-time adaptation cause cognitive overload in self-paced AI tutoring.
Built Maestro with Flutter Web, Riverpod, and GoRouter, powered by a Reinforcement Learning engine with dynamic Z-score scheduling, multi-component reward shaping, local Markdown RAG via LLM APIs, and iOS Safari memory cap handling.
Maestro features an adaptive AI tutoring architecture combining a Flutter Web client, Subject Shield context isolator, reinforcement learning policy engine, and real-time Cloud Firestore persistence with local Markdown RAG indexing.