The Epstein File Ranker

Triage the
Record.

There are millions of pages in the Epstein document releases. Court filings, FBI interviews, financial records, flight logs. No single person could read it all. That's why most of it stays buried.

Our AI investigative agent is processing every page. It scores documents for significance, flags names, traces financial connections, and surfaces the leads that matter. All of that feeds into our public data viewer, where you can explore the documents yourself and follow your own leads.

We're also using the same system to write original investigations. When the AI surfaces something significant, we dig in, verify, and publish. Every claim sourced. Every document linked.

Redacted contact-sheet page from DOJ file EFTA00004577.

Contact-Sheet Index

DOJ FTA / EFTA00004577

Redacted scene image from DOJ file EFTA00005538.

Image Exhibit

DOJ FTA / EFTA00005538

Additional collage page from DOJ file EFTA00004577.

Collage Detail

DOJ FTA / EFTA00004577

Raw document page from the DOJ Epstein files corpus.

Raw Document

DOJ FTA / VOL00003

Redacted scrapbook page from DOJ file EFTA00005536.

Scrapbook Cover

DOJ FTA / EFTA00005536

Photo collage page with redactions from DOJ file EFTA00005538.

Photo Collage

DOJ FTA / EFTA00005538

Redacted photo-collage page from DOJ file EFTA00004577.

Collage Segment

DOJ FTA / EFTA00004577

Before and after comparison page from DOJ file EFTA00005538.

Before / After

DOJ FTA / EFTA00005538

Third page from DOJ file EFTA00005538.

Exhibit Page

DOJ FTA / EFTA00005538

Redacted evidence photo from DOJ file EFTA00005567.

Evidence Photo

DOJ FTA / EFTA00005567

Multi-person photo collage from DOJ file EFTA00005538.

Social Collage

DOJ FTA / EFTA00005538

Raw document page from the DOJ Epstein files corpus.

Correspondence

DOJ FTA / VOL00003

How It Works

An AI That Reads. Tools You Can Use. Journalism That Matters.

The scale of these releases is designed to overwhelm. We built a system that refuses to be overwhelmed.

AI Processes Everything

Our agent reads millions of pages the way a human would: text, layout, redactions, handwritten notes. Every document gets scored for investigative significance and tagged with names, dates, and connections.

You Explore the Data

The scored documents feed into our public data viewer. Filter by significance, search by name, drill into original DOJ files. Find leads we might have missed. The tools are yours to use.

We Investigate the Leads

When the AI surfaces something significant, we dig in. Verify the connections. Follow the paper trail. Publish what we find with full source citations so you can check our work.

Coming Soon

Submit a Lead

Found something in the documents? Have a name or connection you want us to investigate? Soon you'll be able to submit leads directly to our AI investigator. It will analyze the relevant files, surface the connections, and we'll publish what we find.