Advanced Free 7 hours

Advanced RAG Engineering

Hybrid search, reranking, evaluation, and retrieval-quality engineering.

A rigorous advanced course on retrieval quality engineering through hybrid search, reranking, advanced indexing patterns, evaluation metrics, and systematic failure analysis. The curriculum is grounded in the curated technical references and preserves their technical depth, code, diagrams, decision rules, and failure analysis. Every lesson adds objectives, an applied lab, review questions, and concrete completion evidence. The course closes with a substantial capstone: Run a measured RAG improvement study comparing chunking, hybrid retrieval, reranking, and prompting against a fixed evaluation set.

What you will learn

Leave with a working mental model.

  1. 01 Diagnose whether failures originate in ingestion, retrieval, ranking, or generation
  2. 02 Combine lexical and semantic retrieval with calibrated reranking
  3. 03 Design evaluation datasets and measure retrieval and answer quality
  4. 04 Apply advanced RAG patterns without hiding complexity behind framework defaults

Before you begin

Prerequisites

  • A working basic RAG pipeline
  • Embeddings and vector-search knowledge
  • Comfort evaluating model outputs

Curriculum

Course contents

7 modules · 13 lessons

01 Module 1: Failure-Oriented Architecture 2 lessons
02 Module 2: Advanced Chunking and Representation 2 lessons
  1. 01 RAG Chunking Strategies
  2. 02 RAG Embeddings and Models
03 Module 3: Hybrid Retrieval 2 lessons
  1. 01 RAG Retrieval Strategies
  2. 02 RAG Hybrid Search
04 Module 4: Reranking and Context Selection 1 lesson
  1. 01 RAG Reranking
05 Module 5: Evaluation 1 lesson
  1. 01 RAG Evaluation Metrics
06 Module 6: Advanced Patterns and Applications 3 lessons
  1. 01 RAG Advanced Patterns
  2. 02 RAG Generation and Prompting
  3. 03 RAG Real-World Use Cases
07 Module 7: Capstone Studio 2 lessons
  1. 01 Capstone Brief and Architecture
  2. 02 Validation, Review, and Definition of Done