Appearance
Document Processing
Elsai provides dedicated packages for extracting content from every common document format.
Packages
| Package | Version | Purpose |
|---|---|---|
elsai-text-extractors | 0.1.0 | PDF, DOCX, CSV, Excel — structured text extraction |
elsai-ocr-extractors | 2.0.1 | OCR from scanned PDFs — Azure, Amazon Textract, Mistral, LlamaParse |
elsai-parsers | 0.1.0 | Natural language queries over Excel/CSV using an LLM |
elsai-nli | 0.1.0 | Natural language interface for CSV data |
Choosing the right package
Document type Recommended package
─────────────────────────────────────────────────────
Digital PDF/DOCX elsai-text-extractors
Scanned PDF/image elsai-ocr-extractors
Excel (NL query) elsai-parsers
CSV (NL query) elsai-nli