Sheaf

We read the fine print.
We catch what they skipped.

How extraction works
Scrutiny Engine / Extraction Data Hover to reveal structure

A 104-page packet.
Every edge found.

How splitting works
Boundary Engine / 104 pp. → 33 documents Hover to run the packet

Every file, against
every requirement.

How matching works
Match Engine / 16 requirements · 11 covered Hover to test coverage

Identity2 / 2

Government photo ID Driver Licensepp. 1–2 Approved
Borrower identity affidavit Signature Name Affidavitp. 3 Approved

Income2 / 4

Income, tax year 2022 1040 + Schedule Cpp. 4–19 Approved
Employment verification Pay Stub ×2pp. 20–23 Recommended
W-2, most recent year W-2, tax year 2021pp. 24–25 Superseded
Written verification of employment no document found Missing

Assets2 / 3

Asset statement, 60 days Bank Statement ×2pp. 26–33 Approved
Large deposit explanation Letter of Explanationp. 34 Recommended
Gift letter, donor signed no document found Missing

Collateral3 / 4

Executed purchase agreement Purchase Agreementpp. 35–48 Approved
Appraisal, full interior Appraisal Reportpp. 49–71 Pending
Title commitment Title Commitmentpp. 72–83 Approved
Flood determination Flood Certificatep. 84 Approved

Compliance2 / 3

Hazard insurance binder no document found Missing
Notice of right to cancel Notice of Right to Cancelpp. 85–87 Approved
Taxpayer consent, 4506-C Form 4506-Cp. 88 Recommended

Arrived, not required2

Nothing on the checklist asks for this Utility Billp. 89 Unassigned
Filed under a different application Prior Loan Estimatepp. 90–92 Unassigned

Correct it once.
It never comes back.

How review works
Review Desk / learned from 1,140 decisions Hover to apply the learning

Learned from your reviewers5 rules

Prior-year W-2 offered against a current-year requirement Request the current yearapplied 212× Automatic
Bank statement missing its final summary page Accept if the balance page is thereapplied 96× Automatic
Unlabelled scan that reads as a title commitment File as title, not as an exhibitapplied 341× Automatic
Gift letter arriving without a donor signature Hold pending signatureseen once Learning
Two appraisals, same property, different dates Keep the later oneseen once Learning

Always yours to sign3 rules

Any file above the $2M authority line Named underwriternever automatic Reserved
Exceptions to the flood-zone requirement Named underwriternever automatic Reserved
First file from a new correspondent Named underwriternever automatic Reserved
Sheaf / Test Environment Close

Drop your files here

or click to choose

Multiple PDF, PNG, JPEG & TIFF accepted

Selecting files opens your OS picker. Processing continues in Sheaf. No files are uploaded from this page.

Engineered for the document scale of institutions such as (illustrative examples):

Cadence Bank Old National Bank Webster Bank Hanover Insurance Cincinnati Financial Stewart Title
01 / INPUT

Bring the mess.

Do not prep. Do not sort. Do not clean up. Just point Sheaf at the raw intake and let the engine handle the reality of operational documents.

Scans
Phone Photos
Merged Packets
Application Dumps
Wrong-order Pages
Junk Pages
Dense Tables
Signatures
02 / PROCESS

The document work.

What Sheaf does while you aren't looking. A complete forensic pipeline that turns a dumped pile into a structured dossier.

Finding the edges.

One file arrives holding thirty documents and no table of contents. Sheaf decides where each one starts and ends, and covers every page while doing it.

Closing the checklist.

Your policy says what an application needs. Sheaf binds what actually arrived to what was actually required, and names the difference.

Learning your desk.

Every correction your reviewers make is a decision Sheaf keeps, and then applies. The queue gets shorter each week, and the rules it runs on are yours rather than an industry average.

Separate

Find the edges.

Detects exactly where one document ends and the next begins inside a massive, unstructured merged packet.

Name

Call it what it is.

Recognizes the document context by its content, entirely ignoring useless file names or absent metadata.

Classify

Put it in place.

Groups the newly separated and named documents into logical, policy-driven dossiers.

Extract

Pull the data.

Reads the fine print, the tables, and the handwritten notes to pull exactly the fields your operation requires.

Review

Surface the human bits.

Never pretends it is certain. Low-confidence extractions are flagged for an operator's final judgment.

Read

Every page, first.

The whole packet goes through text and layout analysis before anything is cut, so scans and photographs count the same as clean digital pages.

Signal

Find the seams.

Looks for the marks a new document leaves behind: a fresh header, a restarted footer, a "page 1 of 6" counter, a change of form entirely.

Cut

Commit the boundary.

Places every cut so the runs tile the packet end to end. No page is left sitting between two documents, and none is claimed by both.

Name

Label the run.

Each separated document is typed from its own content, so a run named scan_0042.pdf still comes out as a title commitment.

Require

Read the checklist.

Your policy is the list of things that must be satisfied. It varies by product, by program, and by applicant, and Sheaf takes it as given.

Candidate

Line up the files.

Everything that arrived is a candidate, including the documents nobody labelled and the ones filed under the wrong application.

Bind

Satisfy, or don't.

Binds each document to the requirement it actually satisfies, with a confidence attached, and abstains rather than guessing when the evidence is thin.

Gap

Name what's missing.

Reports the requirements nothing covered. The chase list writes itself, and it is the same list your auditor would have built.

Queue

Only what needs you.

Work reaches a person for one of two reasons: the evidence fell short of your threshold, or your policy puts a signature on this class of decision. Everything else is already filed.

Decide

One call, fully recorded.

Each decision keeps the actor, the pages it covered, and what the engine had proposed before anyone touched it — so the disagreements are visible, not just the outcomes.

Learn

The correction becomes a rule.

When the same situation resolves the same way twice, it stops reaching the queue. Sheaf applies the decision from then on, and cites the two that created it.

Diverge

Your shop, not the average.

What is learned here is scoped to your organization. Nothing your reviewers decide leaves it, and nothing from anyone else’s desk arrives in yours.

Technical detail How extraction works, and why a value has to point at a page Read the spec →
Technical detail How splitting works, and what it costs to lose a page Read the spec →
Technical detail How matching works, and where extraction joins it Read the spec →
Technical detail How review works, and what it takes to promote a decision to a rule Read the spec →
03 / OUTPUT

Structured payload.

The file doesn't need to arrive perfect, but the data leaves that way. Ready for your existing infrastructure.

Extracted data is formatted, normalized, and deposited directly into your existing databases, loan origination software, or systems of record.

No new portals to check. Just accurate data, exactly where you expect it to be.

{
  "dossier_id": "req_8849-2A",
  "document": {
    "type": "1040_schedule_c",
    "confidence": 0.99
  },
  "extracted": {
    "borrower_name": "Samuel Hayden",
    "adjusted_gross": 142500,
    "requires_review": false
  }
}

One packet in, a list of real documents out — each with the page range it occupies and the type it was recognized as.

The ranges tile the original file end to end, so nothing you were sent goes missing on the way through.

{
  "packet_id": "pkt_8849-2A",
  "pages_in": 104,
  "documents": [
    { "type": "loan_application", "pages": "1-4" },
    { "type": "pay_stub", "pages": "5-6" }
  ],
  "documents_out": 32,
  "pages_orphaned": 0
}

Coverage against your checklist, requirement by requirement, with the document that satisfied each one and the confidence behind it.

What's outstanding comes back as a list you can act on, not a percentage you have to interpret.

{
  "checklist_id": "chk_8849-2A",
  "satisfied": 6,
  "outstanding": [
    { "requirement": "hazard_insurance" },
    { "requirement": "gift_letter" }
  ],
  "complete": false
}

The queue comes back as work, not as a dashboard. What needed a person, what a learned rule already settled, and which decisions it was learned from.

Every automatic resolution points back at the human ones behind it, so an auditor can walk from the outcome to the people without leaving the record.

{
  "review_id": "rev_8849-2A",
  "queued_for_human": 2,
  "settled_by_rule": 11,
  "rules": [
    { "id": "prior_year_w2", "learned_from": 2 }
  ],
  "named_signer_required": true
}

What document work actually requires

Page-level citations

Every extracted value carries the file, page and span it came from. An answer you cannot walk back to the page is an answer you cannot defend.

Scanned and photographed documents

Real work arrives as a scan of a print of a fax. A page with no text layer is the ordinary case here, not the exception that gets routed to a person.

Knowing what is missing

Completeness is the question most systems never answer. Sheaf names the requirement that nothing satisfied, which is usually what is holding the file up.

Audit trail and reviewer decisions

Every automatic outcome points back at the human decisions behind it, so an auditor can walk from a value to a page without leaving the record.

Security and data handling

Every request is checked against the role that made it, records age out on a fixed retention window, and each value stays traceable to its page.

Compare