RAG with Python Cookbook

Chapter 4. Retrieval

A Note for Early Release Readers

With Early Release ebooks, you get books in their earliest form—the author’s raw and unedited content as they write—so you can take advantage of these technologies long before the official release of these titles.

This will be the 7th chapter of the final book. Please note that the GitHub repo will be made active later on.

If you have comments about how we might improve the content and/or examples in this book, or if you notice missing material within this chapter, please reach out to the author at [e-mail].

Improving the retrieval step is often one of the most effective ways to boost the overall performance of a RAG system.

There is no single right or wrong approach here. As with all other machine learning use cases, you need a solid toolkit to be creative. Often, the simplest solutions are most effective.

Figure 4-1 shows several techniques you can try when fine-tuning the retrieval step of your RAG system. For each of them, you can find suitable recipes in this chapter:

Table 4-1. Retrieval enhancement techniques

Technique	Description	Recipe Reference
Metadata Filtering	Using metadata to filter search results based on what you know about the user	Recipe 4.1
Multi-Query Retrieval	Creating multiple versions of the same prompt to find more relevant documents	Recipe 4.3
Query Routing System	Using a query routing system to identify the best data sources or tools to use to answer the question	[Link ...