Friendly Bytes
All training courses
AI Engineering

RAG & Retrieval

RAG and Retrieval for Knowledge-Based AI Systems

Reliably connect your own knowledge sources with Large Language Models.

Training format

2–4 days, depending on the depth and focus areas.

Content and duration are tailored to the team and the desired focus areas.

Area

AI Engineering

Duration

2–4 days, depending on the depth and focus areas.

Approach

Practical

Language

German or English

01

About the training

The training approaches Retrieval-Augmented Generation as an engineering problem. From document preparation and chunking to embeddings, traditional search, retrieval strategies, and evaluation, it builds an understanding of how robust knowledge-based AI systems are designed.
02

Learning objectives

Understand and design RAG architectures Prepare documents effectively for retrieval Assess embedding and retrieval methods Systematically improve retrieval quality Evaluate RAG systems and analyze common failure modes
03

Content and focus areas

RAG architecture Ingestion and document preparation Chunking strategies Embeddings Vector search BM25 and traditional search Hybrid retrieval Query transformation Reranking Context building Evaluation and observability Security and access control
04

Practical work in the training

Participants build a retrieval pipeline step by step and examine how chunking, search methods, queries, and reranking affect the quality of the provided context.
05

Tailored training plan

The training is tailored to your team.

The content described provides the technical framework. The topics we explore in depth, the examples we use and the time we devote to each area depend on the participants' existing knowledge, goals and technical challenges.

Discuss your training needs