Best Practices

  • Upload clean, well-structured documents.
  • Keep information up to date; re-ingest after significant changes.
  • Use descriptive File Name values.
  • Organize with consistent Category and Tags.
  • Remove duplicate or obsolete files.
  • Monitor Failed ingest jobs regularly.
  • Prefer Default (Docling decides) chunking unless you have a specific token strategy.
  • Verify DNS for larger website crawls when page credits require it.