Advanced RAG Techniques Guide

Lessons

Modules

🎓

Bootcamp access

Lo que aprenderás

✓Design and implement a complete RAG pipeline (indexing, retrieval, generation, evaluation)

✓Apply chunking strategies (fixed, semantic, recursive) and select the optimal one for your data

✓Optimize user queries with expansion, rewriting, decomposition, and HyDE techniques

✓Implement re-ranking with cross-encoders, LLM-based re-rankers, and Cohere Rerank

✓Build hybrid search combining BM25 keyword search with semantic embeddings

✓Use metadata filtering for multi-tenant isolation, time-based, and category-based retrieval

✓Deploy RAG to production with Pinecone (namespaces, scaling, migration from ChromaDB)

✓Evaluate RAG quality with RAGAS framework, golden datasets, and regression testing in CI/CD

¿Para quién es?

•AI Engineers who have built basic RAG systems and need production-level precision and scalability
•Backend developers building AI-powered search, Q&A, or document intelligence features
•Engineers whose RAG prototypes work in demos but fail with real user queries
•Teams optimizing existing RAG systems that return irrelevant results or scale poorly
•Developers preparing for AI engineering roles that require advanced retrieval expertise

Requisitos

•Python intermediate (OOP, async/await, type hints)
•Basic RAG implemented (embeddings, vector search, LLM generation)
•Experience with ChromaDB or similar vector database
•Familiarity with OpenAI API or equivalent LLM provider
•Understanding of embeddings and cosine similarity concepts

Course content

1Módulo 1: RAG Pipeline Completo (Architecture Overview)8 lessons

2Módulo 2: Chunking Strategies8 lessons

3Módulo 3: Query Optimization8 lessons

4Módulo 4: Re-ranking Techniques8 lessons

5Módulo 5: Hybrid Search (BM25 + Semantic)8 lessons

6Módulo 6: Metadata Filtering8 lessons

7Módulo 7: Production RAG con Pinecone8 lessons

8Módulo 8: RAG Evaluation + Proyecto Integrador8 lessons

Reviews

What students say

These reviews are from enrolled students who completed at least 50% of the course. We moderate reviews only on content grounds (spam, offensive language, personal data), never for being critical or negative.

No approved reviews yet.

Be the first to share your experience!