Upload PDF or DOCX
Upload your product manuals, policy documents, or knowledge base files and have your bot trained on them in minutes.
Supported formats
DGbot accepts the following file types:
| Format | Extension | Max file size |
|---|---|---|
.pdf | 50 MB | |
| Word document | .docx | 20 MB |
| Plain text | .txt | 5 MB |
| JSON (structured) | .json | 10 MB |
Scanned PDFs (image-only) are not supported. The PDF must contain extractable text.
Steps
Go to Sources in the sidebar.
Open Sources in adminClick Add source and choose PDF / DOCX (or TXT for plain text files).
Select the file type that matches your document.
Click Choose file or drag your file onto the upload area. You can upload up to 5 files at once.
Drag files onto the upload area or click Choose file to browse. Multiple files upload together.
Return-Policy-2024.pdf is easier to manage than doc-final-v3.pdf.Each file processes independently. Large PDFs (100+ pages) may take up to 2 minutes.
All three sources show a Ready badge — training is complete.
Tips for best results
Use text-based PDFs, not scanned images. If you created the PDF from a Word document or design tool, it almost certainly contains extractable text. If it came from a photocopier or scanner, it may not.
Structure matters. PDFs with clear headings, numbered lists, and short paragraphs produce better answers than dense walls of text. If you can, clean up the document before uploading.
Split very large documents. A 500-page product catalog works better as several smaller files grouped by category. The retrieval system can find relevant chunks more precisely.
Upload the freshest version. If you update a document, delete the old source and re-upload the new file. DGbot does not automatically detect file changes.
Troubleshooting
Upload fails immediately — Check the file size limit for your plan and file type. Make sure the file is not password-protected.
Source shows Error badge — Click the error message for details. Common causes: corrupted file, scanned PDF with no extractable text, or a file format that looks like PDF but is actually a different type.
Answers seem wrong or incomplete — Large documents may have relevant content scattered across many pages. Try splitting the document into topic-specific files and re-uploading.