liveBook · Manning

RecursiveCharacterTextSplitter

As explained in Machine Learning Platform Engineering

RecursiveCharacterTextSplitter is a text-splitting utility that recursively divides documents at progressively smaller structural boundaries.

RecursiveCharacterTextSplitter prepares documents for retrieval by splitting oversized text at progressively smaller structural boundaries. It tries headers, paragraphs, sentences, words, and finally individual characters, while overlap preserves context across chunk boundaries. section 12.2.4

RAG indexing configuration

A LangChain pipeline loads Markdown files, splits them with chunk_size=1000 and chunk_overlap=200, then passes the resulting chunks to an embeddings-backed FAISS index.

How Machine Learning Platform Engineering covers it

How deep it goes
Explains recursive boundary selection and provides executable configuration, but defers detailed chunk-quality evaluation.
Examples
Uses Markdown documents, hierarchical separators, 1000-unit chunks, 200-unit overlap, and a test encryption-key query.
What it assumes
Assumes familiarity with documents and retrieval pipelines while demonstrating the splitter configuration from zero.
What you can do after
Apply it via LangChain to create overlapping chunks suitable for embedding and FAISS similarity search.
How it connects
Connects document loading, text chunking, embeddings, vector indexes, and semantic search.
Cover of Machine Learning Platform Engineering
liveBookmanning.com
Read Machine Learning Platform Engineering on liveBook Get the book at manning.com

Also explained in

Compare how 4 books explain RecursiveCharacterTextSplitter →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Machine Learning Platform Engineering. The section links point to the parts of the book it draws on.