An Open-Source Catalog for Learning AI Engineering
Every link below points to a real, freely available course, repo, or reference — no invented handbooks, no fabricated benchmarks. Curated for architects and engineers building on the capabilities mapped elsewhere on this site.
Looking for model leaderboards or the latest model news? See Benchmarks & LeaderboardsStart Learning — Free Courses
Self-paced, no-cost courses from the organizations actually building this stack.
LLM Course
12 chapters covering transformers through fine-tuning — the standard free entry point into modern NLP.
huggingface.co/learn/llm-courseAI Agents Course
5 units plus a capstone covering LangGraph, LlamaIndex and smolagents, with a certification on completion.
huggingface.co/learn/agents-courseGenerative AI for Beginners
21 lessons on generative AI and LLM fundamentals, with runnable code in both Python and TypeScript.
github.com/microsoft/generative-ai-for-beginnersAI Agents for Beginners
18 lessons dedicated specifically to agentic architectures, complementing the generative-AI fundamentals course above.
github.com/microsoft/ai-agents-for-beginnersPractical Deep Learning for Coders
Jeremy Howard's code-first deep learning course — 30+ hours, from your first model through to stable diffusion internals.
course.fast.aiInteractive Prompt Engineering Tutorial
9 chapters plus an appendix on chaining, tool use and retrieval, with a live playground to test every example.
github.com/anthropics/prompt-eng-interactive-tutorialShort Courses
70+ one-to-two-hour modules built with OpenAI, Anthropic and LangChain engineers — prompting, RAG, agents and fine-tuning.
deeplearning.ai/coursesIntroduction to Generative AI
A ~45-minute no-cost path from Google Cloud Skills Boost; pairs with the free Generative AI Fundamentals skill badge.
cloudskillsboost.google/paths/118Open-Source AI Cookbook
Practical, runnable recipes for RAG, evaluation, fine-tuning and agent patterns using open models and tools.
huggingface.co/learn/cookbookOfficial Docs & Reference Repos
Primary-source reference material maintained by the vendors themselves — bookmark these over any third-party summary.
OpenAI Cookbook
Official examples and guides for the OpenAI API — embeddings, fine-tuning, vision, agents
Anthropic Courses
Anthropic's full course repo — prompting, tool use, and real-world application patterns with Claude
Google Cloud generative-ai repo
Sample notebooks and code for Vertex AI and Gemini across RAG, agents, and evaluation
Stay Current — Emerging Research
Where new papers actually land first, before they're written up secondhand.
Vendor AI Certifications
Current as of August 2026 — this is a fast-moving area, several of these have replaced a retired predecessor within the last year, noted below.
Google Cloud
Generative AI Leader
Foundational, non-technical — 50–60 questions, 90 minutes. Aimed at product leads and enterprise architects setting AI strategy, not implementation.
Generative AI Fundamentals (skill badge)
Free, ~2 hours — three short courses culminating in a digital skill badge. The lowest-friction way to start.
Microsoft Azure
AI-103: Azure AI Apps and Agents Developer
The current path. Generative AI and agent workflows are now the largest exam domain. Replaces AI-102, which retires 30 June 2026 — new candidates should go straight to AI-103.
AI-900 → AI-901: Azure AI Fundamentals
Foundational, no prerequisites. AI-900 is being retired and replaced by AI-901 in mid-2026 — the Fundamentals credential itself doesn't expire, only the exam code is changing.
AWS
AWS Certified AI Practitioner (AIF-C01)
Foundational, ~$100. Broad AI/ML/generative AI literacy on AWS — no implementation experience required.
Generative AI Developer – Professional (AIP-C01)
New in 2026, ~$300. Production Bedrock-based GenAI application development, for engineers with 2+ years of cloud experience.
The former AWS Certified Machine Learning – Specialty (MLS-C01) retired 31 March 2026. Its direct successor for MLOps/SageMaker work is the associate-level AWS Certified Machine Learning Engineer – Associate (MLA-C01).