About QuizBuddy Project

This project was inspired by my need to study effectively instead of spending time doing passive learning over and over. It made me research some strategies that can be used and then creating an app to use them.

Efficiency & Learning Strategies

To study efficiently, I digitalized these techniques:

• Active Recall: Testing yourself on concepts rather than passively reviewing notes.
• Spaced Repetition (SM-2): Reviewing concepts at increasing intervals of time.
• The Feynman Technique: Explaining a complex topic in simple terms to a 10-year-old.
• Hard-Start-Jump-to-Easy: Tackling the hardest questions first when focus is highest.
• Brain-Dump Warmup: Writing down everything you know about a topic before starting.
• Pomodoro Focus: A study timer with structured breaks.
• Interleaved Practice: Shuffling questions across different topics.
• Chunking & Mastery: Tracking mastery percentages for small concept chunks in real-time.
• Deliberate Practice: Specifically targeting and practicing weak areas.

Book References

• A Mind For Numbers: How to Excel at Math and Science (Even If You Flunked Algebra) by Barbara Oakley
• Make It Stick: The Science of Successful Learning by Henry L. Roediger III, Mark A. McDaniel, and Peter C. Brown
• Ultralearning: Master Hard Skills, Outsmart the Competition, and Accelerate Your Career by Scott H. Young

---

The Process

OpenCode — Main AI Coding Tool

• Models: GLM-5.2 (plan-mode), DeepSeek-v4-flash (executor).
• Why? GLM-5.2 is a really good model, but it is slow. DeepSeek-v4-flash is cheap, but much faster. This combination is great for cost efficiency and productivity.

Phase-by-Phase Process

• Pre-development: Combination of all the strategies, notes that I read to learn efficiently → Plan out with plan mode to create final spec/product concept → Create one.
• Backend: Based on the documentation → plan out architecture and APIs backend → AI return result, check and approve, refine it if needed → Implement → Create detailed spec for server side.
• Front-end: Based on the server side, plan out page organization and UIs. For page layouts, I asked the model to generate the layout under markdown format to decide the general layout. Then I asked it to implement by giving me different prototypes with different color combinations to look at how it looks.
• Testing: Testing myself, dogfooding using my own product. I asked my friend to use it some time and observe her behavior to improve the experience. Also asked AI to implement test cases.

Thank you for reading, strangerrssss!