10.3 The ingest produced no sections, or almost none

The PDF probably has no extractable text. A scanned document is a picture of words, not words. Run it through OCR and upload the result. If the PDF does have text but the sections came out wrong, the document’s heading structure is what Living Manual reads. A PDF whose headings are manually bolded body text rather than real heading styles gives the extractor nothing to work with. Exporting from the source document with proper heading styles fixes it.