app/services/document_loader.py

from langchain_community.document_loaders import PyPDFLoader
from langchain_text_splitters import RecursiveCharacterTextSplitter

def load_pdf(file_path):

    loader = PyPDFLoader(file_path)

    documents = loader.load()

    splitter = RecursiveCharacterTextSplitter(
        chunk_size=500,
        chunk_overlap=100
    )

    chunks = splitter.split_documents(documents)

    return chunks