Skip to main content
Upload documents to your datasets to make them searchable and available to Stella AI.

​
Supported File Types

FormatExtensionsNotes
PDF.pdfText is extracted from all pages
Word.docx, .docFull document content extracted
Text.txtPlain text files
Markdown.mdMarkdown formatting preserved
File Size Limit: 10 MB per file

​
Upload a Document

Uploading documents requires Admin access.
  1. Navigate to Knowledge Base
  2. Open the dataset where you want to add documents
  3. Click Upload Documents
  4. Either:
    • Drag and drop files into the upload area, or
    • Click Browse to select files from your computer
  5. Click Upload
The document will be queued for processing and appear in your dataset.

​
Document Processing

After upload, documents go through automatic processing:
StatusMeaning
PendingQueued for processing
ProcessingBeing indexed (chunking, embedding generation)
CompletedReady for search
FailedError during processing (check file format)
Processing typically completes within a few minutes. Larger documents may take longer.

​
What Happens During Processing

  1. Text Extraction: Content is extracted from the document
  2. Semantic Chunking: Content is split into meaningful segments (up to 2000 characters each)
  3. Embedding Generation: Each chunk is converted to a vector embedding using AI
  4. Indexing: Embeddings are stored in the database for fast similarity search
Once complete, the document is searchable by Stella AI (if the dataset has “Searchable in Stella” enabled).

​
Delete a Document

  1. Open the dataset containing the document
  2. Find the document in the table
  3. Click the trash icon
  4. Confirm the deletion
Bulk Delete:
  1. Select multiple documents using the checkboxes
  2. Click Delete in the action bar
  3. Confirm the deletion

​
Troubleshooting

Processing usually completes within minutes. If a document stays in “Processing” for an extended period:
  • The document may contain complex formatting
  • Try re-uploading the document
  • Contact support if the issue persists
Common causes:
  • Corrupted file: Try opening the file locally to verify it’s readable
  • Unsupported format: Ensure the file is a supported type (PDF, DOCX, TXT, MD)
  • Password protection: Remove password protection and re-upload
  • Scanned PDF without OCR: PDFs must contain extractable text, not just images
If your file is larger than 10 MB:
  • Split large documents into smaller sections
  • Remove embedded images or unnecessary formatting
  • Compress the file before uploading

​
Best Practices

Use descriptive file names - Document names appear in search results and help users identify content.
  • Keep documents focused on single topics when possible
  • Update documents by deleting the old version and uploading the new one
  • Review document status to ensure processing completed successfully
  • Organize related documents in the same dataset