This repository provides a structured roadmap for aspiring AI Engineers, detailing the necessary skills, learning resources, and tools across three stages: Beginner, Intermediate, and Advanced. It aims to guide individuals from basic LLM API consumption and prompt engineering to building complex RAG applications, agents, and finally, deploying, optimizing, and fine-tuning LLM-powered applications in production (LLMOps).

How It Works

The roadmap is organized into progressive learning stages, each focusing on specific AI engineering concepts and practical applications. It emphasizes a hands-on approach, recommending the development of projects and Proofs-of-Concept (POCs) at each stage. Key areas covered include prompt engineering, working with open-source LLMs, Retrieval Augmented Generation (RAG), vector databases, agentic workflows, and LLMOps.

Quick Start & Requirements

This is a learning roadmap, not a software package. No installation or execution is required. Resources linked within the roadmap may have their own requirements.

Highlighted Details

• Structured Learning Path: Clearly defined stages (Beginner, Intermediate, Advanced) with specific skill sets and learning objectives.
• Resource Curation: Extensive lists of courses, guides, papers, and tools for each skill, including links to platforms like DeepLearning.AI, Hugging Face, LangChain, and specific libraries.
• Project-Based Learning: Practical project ideas are provided for each stage to solidify understanding and build a portfolio.
• Comprehensive Coverage: Addresses core AI engineering topics from basic LLM interaction to advanced deployment and fine-tuning.

Maintenance & Community

The roadmap is maintained by dswh, who actively shares progress and insights on YouTube, X (Twitter), LinkedIn, and Substack. The repository encourages community engagement through starring and following the author's social channels.

Licensing & Compatibility

The repository itself is licensed under the MIT License, allowing for broad use and modification. However, the learning resources and tools referenced within the roadmap are subject to their respective licenses.

Limitations & Caveats

This is a curated guide and not an executable tool. The rapidly evolving nature of AI means some resources or tools mentioned may become outdated. Users are encouraged to verify the current relevance and applicability of the suggested materials.