Earth-OL-Player/ai_learn_project

What it solves

AI development resources are scattered and learning paths are unclear, so developers struggle to get structured practice and feedback. This project bundles a learning roadmap, a hot interview question bank, AI-powered quiz grading with follow-up discussion, a gamified growth system, and a suggestion/comment community into one runnable full-stack application, giving learners a single place to study AI Agent, RAG, tool calling, structured output, and LLM engineering.

How it works

It is a three-tier architecture:

• A Vue 3 + Vite frontend (with TypeScript, Pinia, Vue Router, Element Plus) presents the learning platform, quiz workspace, personal center, and admin pages.
• A Spring Boot backend (Java 17, Spring Security, JWT, Flyway) handles authentication, question bank management, interactions, growth tracking, and admin APIs, backed by MySQL 8.4 LTS.
• A FastAPI AI service (Python 3.11+, Uvicorn) provides AI scoring, discussion, and model capability integration, with a fallback to local rules or external model services.

The user flow: browser -> Vue frontend -> Spring Boot backend -> MySQL for business data, and the backend calls the FastAPI AI service for scoring and discussion. The project also includes a Markdown-based learning route renderer with navigation and a documentation folder for requirements, design, acceptance, and middleware notes.

Who it’s for

Developers learning AI Agent and AI application development, RAG, tool calling, structured output, or LLM engineering who want a concrete, runnable codebase to study and extend. It is also useful for anyone preparing for AI-related interviews or wanting to build a full-stack learning platform.

Highlights

• Curated learning routes and materials rendered as browsable Markdown pages with table of contents and version notes.
• AI-powered quiz workspace: question categories, next-question flow, retry, AI scoring, AI follow-up discussion, and highest-score history.
• Hot interview question bank covering AI fundamentals, Agent basics, RAG full pipeline, vector retrieval, multi-agent, and safety evaluation, with importance scores and real interview counts.
• Growth system with experience points, levels, ranks, badges, and practice records to keep learning motivated.
• Suggestion and comment community with category filtering, hot/newest sorting, and empty-state handling.
• Full-stack closure with Vue 3 frontend, Spring Boot backend, FastAPI AI service, and MySQL database, plus documentation for local startup and acceptance checks.