Agents / RAG

Agentic AI & RAG guide

This is one of the highest-value tracks today. Learn retrieval, chunking, evaluation, tool use, orchestration and failure handling — not just “chatbot” demos.

Best starting path

  • Learn standard RAG before agentic RAG so you understand retrieval basics clearly.
  • Hugging Face courses are useful for agent frameworks and examples.
  • DeepLearning.AI offers stronger structured RAG depth than random YouTube playlists.

Best paid or certification path

  • NVIDIA now includes RAG-oriented preparation within its GenAI ecosystem.
  • DeepLearning.AI’s full RAG course is strong for serious portfolio building.
  • Use platform-specific paths only after you understand generic system design.

Recommended sites

SiteWhy use itBest for
DeepLearning.AI RAGProduction-focused RAG course with architecture, evaluation and deployment ideas.Serious RAG learners
Hugging Face Agents CourseAgentic RAG concepts and practical agent course material.Open-source learners
DeepLearning.AI advanced short courseGreat follow-up for evaluation, groundedness and answer relevance.RAG improvement and eval
NVIDIA GenAI LLM ProfessionalIncludes RAG-oriented learning path items in a broader GenAI credential.Credential seekers

How to save money

  • Use free Hugging Face material first to understand agents.
  • Buy one deep RAG course instead of five shallow short courses.
  • Prioritize courses that teach evaluation, not only vector DB setup.

What to build after learning

  • A PDF Q&A assistant with citations.
  • A campus FAQ bot using retrieval and reranking.
  • An agent workflow that uses search, extraction and answer grading.