Five thousand pages of employment and exposure history, in order.
EEOICPA claim record review reconciles site, job title, era and covered illness against the Site Exposure Matrices, then pulls the AMA Guides 5th findings a Part E impairment rating is built on. Files run 1,500 to 5,000 pages across several productions; every line we return cites the page it came from.
The model is statutory. So is the hearing that tests it.
We do not build for a case type until we can name the standard that defines a correct output and the person who grades the file against it. In EEOICPA Parts B and E, both are written into regulation, down to the statistical model that decides compensability.
Fifteen hundred pages is the light day.
An EEOICPA file runs 1,500 to 5,000 pages, roughly 3,250 in the middle, and it never arrives as one production. Department of Energy employment verification, the DOL case file, dosimetry and incident records, and decades of treating-physician notes land separately, overlap each other, and contradict each other on dates.
What makes the read hard is not length alone. Microfilm-era scans, handwritten shift logs and dosimetry cards, and the same production sent twice under a different cover letter are the normal condition of these files, not the exception.
The volume behind that: DOL logged 9,504 new EEOICPA claims in FY2024, 3,881 under Part B and 5,623 under Part E, served 21,926 claimants, and paid $2.716 billion that year.
Two neighbouring federal reads work the same way: black lung claim review and impairment rating record review.
Scoped on your file, not on a demo.
We do not ship a generic EEOICPA output and hope it fits. You send one real claim file, we read it, and the two of us decide together what the deliverable has to contain before anything is built. The platform organizes, cites, and surfaces; your specialist, your representative, and your rating physician decide.
In the demo case, page 140 turned out to belong to a different claimant. The read flags it and quarantines it on its own line instead of folding it silently into the rest of the record.
The rules the platform never breaks.
HIPAA, under a signed BAA
Every file is handled under our Business Associate Agreement, from the first byte.
Never trains a model
Your records are never used to train any AI model — ours or anyone else's.
Every line cited
If we can't cite it, we don't say it. Every sentence links to its source page.
Deleted after delivery
Files are deleted 30 days after delivery, with a full audit log of every access.
Four capabilities behind every EEOICPA read.
The EEOICPA record review is one deliverable of the same platform that reads, sorts, and cites the whole file.
Medical Chronology
Employment periods, exposure entries and treatment fall into one timeline, synced to every source page.
Handwritten Record Transcription
Shift logs, dosimetry cards and decades-old clinic notes read as text, not skipped as unreadable images.
Missing Records Identification
The gaps in the production are named, so a thin employment period is visible before the hearing rather than during it.
Verifiable AI Citations
If we can't cite it, we don't say it. Every employment period and every finding links to the page it was pulled from.
EEOICPA claim record review, answered.
Send one file. We'll tell you what we can read.
No obligation. If the case type is buildable we'll scope it; if it isn't, we'll say so.