Document Pipeline

Connecting… 2,940 records · 81,724 pages

Scrapers

Czech National Archives Idle

Czech Security Archives Idle

Findbuch Austria Idle

Austrian State Archives Idle

Matricula Online Idle

Manual Import Idle

Ingested

Documents received and queued for processing

0 records · 0 pages

OCR

Reading each page image with AI to extract text and its position on the page

0 records · 0 pages

85,680 / 85,680 jobs 100%

PaddleOCR 0/2 busy

Qwen VL no workers

PDF Build

Building searchable PDFs by overlaying extracted text onto the original scanned images

0 records · 0 pages

5,812 / 5,812 jobs 100%

0/2 busy

Translation

Translating German and Czech text into English using neural machine translation

0 records · 0 pages

170,081 / 170,081 jobs 100%

0/4 busy 58 failed

Embedding

Generating semantic embeddings from translated text for intelligent search

1 records · 289 pages

9,010 / 9,010 jobs 100%

0/1 busy 1 failed

Complete

Fully processed and searchable

2,939 records · 81,435 pages