Skip to main content

Document Production Workspace v1

Know what arrived before deciding what the record supports.

Preserve each production batch, review proposed document boundaries and metadata, inspect duplicate candidates, compare approved documents with reviewer-defined expectations, and export an organized inventory without changing the original evidence.

Exit Protocol organizes and presents document information for professional review. It does not determine legal relevance, discovery sufficiency or a professional conclusion.

The production reality

One matter can arrive through several distinct production paths.

Sorting, separation and organization are necessary professional work. The software is designed to reduce the surrounding mechanical burden while preserving the review itself.

01 / Client-provided

Separate files, inconsistent names

Files may arrive through Drive, Box or Dropbox. One file often represents one document, but names and folder structure may not be useful.

02 / RPD response

One large, mixed PDF

A response may exceed 1,000 pages and combine statements, returns, correspondence and separator pages in request order.

03 / Third-party subpoena

Repeated and supplemental delivery

Similar separation problems can recur across multiple deliveries, with duplicate documents, missing periods and later supplements.

Reviewer-controlled workflow

The software proposes. The professional decides.

01Preserve production batch
02Process source pages
03Review proposed splits
04Approve metadata
05Inspect coverage
06Export inventory
Proposed

Boundary and metadata candidates

Versioned deterministic rules can propose a page range, category, institution, masked account ending, period, tax year or filename.

Needs review

Ambiguous or incomplete evidence

Conflicting dates, formats, account signals, image-only pages and failed extraction remain visible instead of being silently resolved.

Approved

Reviewer-controlled inventory

Only an attributable reviewer action approves document boundaries and metadata for the organized inventory or export.

Historical format changes

Recognition uses semantic evidence, not one fixed bank template.

The current deterministic recognizer looks for combinations of statement terminology, periods, masked account references, page continuity, institution names and government-form identity. It records the engine and profile version with each proposal.

Implemented

Conservative format-drift proposals

  • Older and newer bank and credit-card statement wording.
  • 1040, 1040A, 1040EZ, 1040-SR, 1040-NR and 1040-X identities.
  • Explicit US numeric and English month-name statement periods.
  • Competing or contradictory signals remain unresolved.
  • Image-only pages are never treated as blank separators.
Evidence boundary

Not a production-accuracy claim

The regression corpus contains 21 synthetic cases. Passing that corpus does not establish accuracy on real productions, every institution, degraded scans, every government-form revision or a representative 1,000-page matter.

Representative, authorized and properly de-identified professional examples still require independent evaluation.

Approved work product

An organized directory without renaming the evidence object.

Approved metadata creates a virtual directory and deterministic inventory. The encrypted source object keeps its original identity, hash, filename, production batch and page provenance.

Matter/
├── Tax Returns/
│   ├── 2022 — Married Filing Jointly.pdf
│   └── 2023 — Married Filing Separately.pdf
└── Bank of America/
    ├── Checking — Account ending 4821/
    │   ├── 2024-01 Statement.pdf
    │   ├── 2024-02 Statement.pdf
    │   └── 2024-04 Statement.pdf  [first seen: Batch 4]
    └── Needs Review/
        └── unresolved-source-pages.pdf

Expectations and apparent gaps

The professional defines what should be present.

Exit Protocol does not guess that a document is missing. It compares approved documents only with an expectation explicitly created by a reviewer.

ExpectationAfter Batch 2After Batch 4Meaning
Checking ending 4821 · April 2024 Apparent gap Available First seen in Batch 4. Availability does not determine sufficiency.

Professional boundary

What the workspace does not decide.

Legal relevance

No automatic relevance, responsiveness or legal-significance determination.

Discovery sufficiency

No conclusion that a production is complete or sufficient for an assignment.

Professional conclusion

No opinion about ownership, tracing method, admissibility or the meaning of a missing document.

Professional evaluation

Challenge the workflow with synthetic or properly de-identified records.

The useful outcome is identifying where the proposed organization becomes professionally wrong or impractical. Do not submit client documents through the public inquiry form.