Generative AI and Agentic AI: A Treasure Trove of Learning Resources from Theory to Practice

This post introduces the GitHub project GenerativeAI-and-Agentic-AI maintained by Mrunmay07. It is a systematic learning guide for Generative AI and Agentic AI, covering core technologies like large language models (LLMs), prompt engineering, RAG retrieval augmentation, AI agent architectures, tool calling, memory systems, and workflow orchestration—suitable for developers from beginners to advanced levels. Original project link: https://github.com/Mrunmay07/GenerativeAI-and-Agentic-AI, published on 2026-05-31.

In today's rapidly developing AI landscape, Generative AI and Agentic AI have become hot topics, but developers lack a systematic learning path. This project addresses this pain point: it is not just a list of links but a carefully organized knowledge base covering a complete learning path from basic concepts to advanced applications, suitable for both novices and experienced developers.

The project focuses on Generative AI and Agentic AI:

1. Large Language Models (LLMs)：Introduces the principles and capability boundaries of mainstream models (GPT, Claude, Llama, etc.), as well as advanced topics like fine-tuning and quantization;
2. Prompt Engineering: Explains prompt design principles (role setting, context management, few-shot learning, etc.), chain-of-thought reasoning techniques, and reusable template construction;
3. RAG (Retrieval-Augmented Generation): Explains the RAG architecture (combining external knowledge bases with generative models), vector database selection, embedding model comparison, retrieval strategy optimization, etc.;
4. AI Agent Architectures: Introduces mainstream architectures like ReAct and Reflexion, enabling AI's autonomous planning, tool calling, and collaboration capabilities;
5. Tool Calling and Memory Systems: Explains tool calling mechanisms (function calls, API integration, error handling) and memory systems (conversation history, user preference management);
6. Workflow Orchestration: Introduces frameworks like LangChain and AutoGen to build complex automated workflows for multi-agent systems.

The project is strongly practice-oriented, with code examples and cases (e.g., RAG question-answering systems, tool calling examples) for each topic. Application scenarios include:

• Enterprise knowledge management: Internal document Q&A based on RAG;
• Automated customer service: Multi-turn dialogue agents;
• Code assistance development: Tool-calling programming assistants;
• Data analysis: Automatic querying and report generation;
• Content creation: Generative AI and manual review workflows.

Beginners are advised to follow this order: Basic prompt engineering → RAG technology → Tool calling → Agent architectures and workflow orchestration; Experienced developers can directly dive into topics of interest. The resource library provides advanced materials and best practices to help build production-level AI systems.

Generative AI and Agentic AI are reshaping the way software is developed. This resource library provides a systematic entry point for learning, covering a knowledge system from theory to practice, and is continuously updated to keep up with the latest technologies. For developers, mastering these technologies is a must to stay competitive in the AI era, and this resource library is an ideal starting point.