AI Learning Management System. This project involved reconstructing a legacy system to create a cutting-edge AI-powered Teaching SaaS for international schools. Key achievements include a zero-downtime migration and supporting four R&D lines with a single design resource.
Exam Assembly Workflow Reconstruction
Problem: The original process was lengthy and complex.
Solution: We introduced a "Shopping Cart" logic for question selection, allowing users to "add to cart" and preview dynamically.
Result: Average assembly time dropped from 15 minutes to 5 minutes (60%+ efficiency gain). Users reported they are "finally not afraid of assembling exams".
Exam Assembly Workflow
AI Grading & Implementation
Efficiency First: We designed with "Negative Operation Logic". Assuming AI accuracy is high, teachers only need to intervene for errors rather than confirming every correct suggestion.
Result: Grading time reduced from 20 minutes to 5 minutes (80% efficiency gain) with a 90% AI adoption rate.
AI Grading Interface
Design Ops & Scalability
1 vs 4 Leverage: Supported 4 development teams with a single design system.
1-to-N Abstraction: Abstracted the core "Assign-Take-Grade" architecture to easily replicate success from TOEFL to IELTS, achieving >70% configuration reuse.
Agile Loop: Established a direct feedback loop for rapid iteration.
Design Ops Architecture