Strongly recommend using Apache Tika[1] for this. It's industry standard for ubiquitous document text extraction. You can take the text output from Tika, chunk

Strongly recommend using Apache Tika[1] for this. It's industry standard for ubiquitous document text extraction. You can take the text output from Tika, chunk it with something like Chonkie[2], and embed it for your search index. -[1]https://tika.apache.org/ https://tika.apache.org/ -[2]https://chonkie.ai/ https://chonkie.ai/