Chunking Documents for RAG | Dataford Interview Questions

Write a script to process and chunk large documents for a RAG pipeline.

Implement chunk_documents(documents, max_words, overlap_words), which accepts a list of documents containing id and text, then returns ordered chunk dictionaries containing id, chunk_index, and whitespace-normalized text. Each chunk may contain at most max_words words, consecutive chunks from the same document must share overlap_words words, and empty documents produce no chunks.

Constraints

1 <= len(documents) <= 100
0 <= total number of whitespace-separated words <= 1000
1 <= max_words <= 1000
0 <= overlap_words < max_words
Each document contains an id and a string text
Return chunks in input document order

Function Signature

def chunk_documents(documents, max_words, overlap_words):

Your question is Chunking Documents for RAG. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.

Problem

Write a script to process and chunk large documents for a RAG pipeline.

Constraints

1 <= len(documents) <= 100
0 <= total number of whitespace-separated words <= 1000
1 <= max_words <= 1000
0 <= overlap_words < max_words
Each document contains an id and a string text
Return chunks in input document order

Function Signature

def chunk_documents(documents, max_words, overlap_words):

Your question is Chunking Documents for RAG. Start with the requirements in the Question tab.

Run and submit as often as you like. When you're ready, talk me through your approach or go straight to the code.