Uploading documents — an overview

Learn what file types Vulgate accepts, how ingestion works, and how to track progress.

May 21, 2026

Uploading is the first step in the Vulgate pipeline. Once a file is uploaded, our ingestion process turns it into structured, searchable, AI-ready content — without you having to lift a finger.

The lifecycle of a document

Every file you add goes through the same sequence:

  1. Upload. The file is securely transferred to Vulgate. Large files are split into parts and uploaded in parallel.
  2. OCR & text extraction. For PDFs, images, and scans, Vulgate uses computer vision and OCR to extract clean text. Audio files go through speech-to-text.
  3. Structuring. The raw text is segmented into chapters, sections, paragraphs, and footnotes. The result is encoded in TEI XML — the gold-standard format for scholarly editions.
  4. Embedding. Every section is converted into a vector so neural search can match meaning, not just keywords.
  5. Publication. When ingestion completes, the document is visible in your Library, ready for search, chat, and translation.

How to upload

  1. From the sidebar, open Uploads.
  2. Click Upload, then drag-and-drop files into the upload area or pick them from disk.
  3. Confirm to start the upload.

Files start uploading immediately and ingestion begins automatically.

Where to watch progress

Open Uploads → Processing to see every running, queued, or recently finished ingest job. Each row shows the document title and current status. Click into a job to see per-step progress and a percentage. You can leave the page — ingestion continues in the background.

You'll also receive an in-app notification when a job completes or fails.

File size and quota

Single files are capped at 1 GB. Large files are uploaded in parts (multipart upload) automatically. Your plan determines the total storage and number of ingest credits available per month — each page consumes credits according to the processing tier you choose. Check Settings → Billing for your current allowance.

What to do next

Search help