Uploading PDF Documents
Learn how to upload PDF files to train your chatbot. Searj analyzes document content and uses it to accurately answer your customers' questions.
Supported Formats
📄 Searj supports PDF files only. Maximum file size is 10MB per file.
How to Upload a PDF
Follow these steps to upload a PDF to your bot:
From the dashboard, select the bot you want to train.
Click on the "Documents" tab on the bot page.
Click the "Upload PDF" button and select the file from your device.
Searj will process the file automatically. The document status will update once complete.
Processing Pipeline
After uploading a PDF, it goes through several stages before becoming ready for use:
The file is securely stored on Searj servers (MinIO) with full encryption.
Text is extracted from all pages of the document.
Content is split into smaller chunks to optimize search accuracy.
Numerical representations (embeddings) are created for advanced semantic search.
The document is now ready and the bot can use it to answer questions.
Document Status Indicators
Track each document's progress through these status indicators:
| Status | Icon | Description |
|---|---|---|
| Pending | ⏳ | File has been uploaded and is waiting to be processed. |
| Processing | 🔄 | The file is currently being analyzed and content is being extracted. |
| Completed | ✅ | Document is ready and the bot can use this content. |
| Failed | ❌ | An error occurred during processing. Try re-uploading. |
Tips for Best Results
- Use text-based PDFs — Files created from programs like Word work excellently. Scanned image PDFs may not extract well.
- Arabic PDFs work great — Searj fully supports Arabic content and accurately understands Arabic text.
- Keep PDFs organized — Use one file per topic for more accurate answers.
- Larger files take longer — Larger files take more time to process. Be patient with big documents.
Plan Limits
Maximum content size varies by your plan:
| Plan | Content Size |
|---|---|
| Free | 400 KB |
| Starter | 10 MB |
| Pro | 20 MB |
| Business | 40 MB |
| Enterprise | 100 MB |
Reprocessing a Document
If a document failed to process or you want to refresh the embeddings, you can reprocess it:
Locate the document you want to reprocess.
Click the reprocess button next to the document. Processing will start from scratch.
💡 Reprocessing is useful when AI models are updated or when initial processing failed.
Deleting Documents
To delete a document, go to the Documents tab and click the delete icon next to the document you want to remove.
⚠️ When you delete a document, all associated embeddings are removed. The bot will no longer be able to answer questions about that document's content. This action cannot be undone.