Keywords
Game-based Learning, AI, Prompt Engineering

Technologies
Next.js, Phaser, Large Lanuage Model, SQLite

This project was developed as a Serious Game which aims to help teach and practice Python programming with engaging mechanics.

The game incorporates Pedagocial Approaches such as Retrieval practice, Elaboration, Concrete examples, Dual-coding, Interleaving and Spaced practice.

All of the educational content is generated using a Large Langauge Model (LLM). This approach allows for diversity in problems for players of similar skill level, allowing replayability. This also removes the limit for how difficult problems can be.

An adaptive skill algorithm is used to effectively generate problems suitable for the skill level of the player. This allows for the generated problems to not be too difficult or too easy for a player.

Players can view a history of all problems they have answered, allowing for the user to learn from their mistakes. If assistance is required, the player can request a detailed AI explanation regarding the solution to a previous problem.

The game also includes player incentives such as daily challenges and a shop where players can purchase crates which contain items, providing in-game boosts.