gitmyhub

AgentGuide

HTML ★ 6.3k updated 17d ago

https://adongwanai.github.io/AgentGuide | AI Agent开发指南 | LangGraph实战 | 高级RAG | 转行大模型 | 大模型面试 | 算法工程师 | 面试题库 | 强化学习|数据合成

A Chinese-language structured study guide and job-search roadmap for developers wanting to learn AI agent development using frameworks like LangChain, AutoGen, and CrewAI.

PythonLangChainLangGraphAutoGenCrewAIMilvusChromasetup: moderatecomplexity 3/5

AgentGuide is a Chinese-language study guide aimed at developers who want to learn AI agent development and land a job in that field. The README describes it as a structured, job-search-oriented resource rather than a simple link dump. The content is written primarily in Chinese, with some English technical terms throughout.

The guide covers several technical areas: building AI agents using frameworks like LangChain, LangGraph, AutoGen, CrewAI, and Swarm; building RAG (retrieval-augmented generation) systems that let language models answer questions using external documents; setting up vector databases such as Milvus and Chroma; and training or fine-tuning language models using techniques like LoRA, reinforcement learning from human feedback, and similar approaches. It also covers multi-agent systems where multiple AI components collaborate on a task.

Alongside the technical content, the guide includes a job search roadmap. This covers how to choose between an algorithm-research role and an engineering-development role, how to frame your resume around what you built rather than what you studied, how to find hiring managers on LinkedIn and reach out directly, and how to approach the interview process for AI-focused positions. The guide describes a six-step path from deciding on a target job type all the way to preparing for offers.

The project also curates links to open-source tools, practice projects, and a question bank of over 1,000 interview questions. The author describes themselves as a working large-model algorithm engineer, and positions the guide as distilling practical field experience rather than academic theory.

The full README is longer than what was shown.

Where it fits