Document Upload

Document upload is sending a file to DocLD for processing. After upload, parsing extracts text, tables, and layout; OCR runs for scanned or image content; chunking segments the content; and embedding stores vectors in the vector index for vector search and RAG.

Flow

1. Upload — File is sent via API or dashboard
2. Job — A job is created; processing runs (often async processing)
3. Complete — Webhook or polling notifies when the source document is indexed and ready for RAG and extraction

Supported Formats

See file format for supported extensions (PDF, images, spreadsheets, documents). Batch processing supports uploading multiple documents in one operation.

Related Concepts

Document upload is the entry point for source documents. Parsing and OCR process uploaded files. Jobs and webhooks track completion. Knowledge bases organize uploaded documents.