Autor: MedaBeda ++++
Datum: 13. 02. 2026 14:49:18
https://epsteinexposed.com/
https://old.reddit.com/r/Epstein/comments/1r3joqr/i_mapped_every_connection_in_the_epstein_files_it/
A week ago I posted about an open database I’ve been building to cross reference Epstein case material. That post did way better than I expected (568k views, 4.6k upvotes) and it hugged my server to death twice.
Since then I basically did nothing but ingest, clean, and index more data. The database is now big enough that “just read the docs” is not advice, it’s a cry for help.
What it was last week
~6,000 documents
1,708 flights
2,700 emails
1,438 people
What it is now
1,522,060 documents (all DOJ releases we have access to so far), full text searchable
1,708 flights (1997 to 2019) with manifests where available
10,000+ emails indexed with threading
1,350 people (cleaned: removed duplicates + nuked a bunch of false connections)
638,000 docs run through redaction analysis
~1.8M individual redactions detected
~616k flagged by our tooling as “looks questionable, take a closer look”
~39,500 pages of text recovered from under black bars (you can see examples on the site)
107,000 named entities pulled out via NLP (people, orgs, places, dates)
1,530 audio/video transcripts
4,300+ photos/media (raid photos, exhibits, property shots, government releases)
That’s not a typo: 1.5 million documents.
If you search a phrase, it searches inside the actual pages (OCR where needed) and email bodies, not just titles. |
|