The problem
Post-close teams receive large, unsorted closing package PDFs and must manually split them into individual documents, verify signatures and notary information page by page, track condition clearance against investor-specific checklists, and prepare an audit-ready file for delivery. This takes 45–60 minutes per loan and produces throughput that is 4× lower than what automation can achieve.
The investor-delivery risk is real: a missed condition or incorrect document version causes a buyback. Manual review at scale means some percentage of loans always slip through with unresolved issues — not because reviewers are careless, but because the volume makes exhaustive review impossible.
| 45–60 min of manual package review per loan Splitting large PDFs, stacking documents into the right folder structure, verifying signatures and dates page by page — all performed manually before a single exception can be identified. |
Missed conditions cause buybacks Investor-specific validation rules are checked manually against paper or spreadsheet checklists. At volume, conditions get missed. Buybacks are expensive and avoidable. |
Manual checklist tracking per investor Each investor has different requirements for document presence, signature placement, notary completeness, and date constraints. Maintaining and applying those rules by hand is error-prone and does not scale. |
How it works
The Processing & Closing workflow uses the same source-file upload endpoint as Loan Setup. Your integration uploads the closing package PDF; MOS runs the full pipeline — split, classify, extract, validate, export — and surfaces only the exceptions that need human review.
| Step 1 — Upload the closing package PDF POST the package to /loans/{loanId}/source-files as multipart/form-data. Packages of 100–300 pages are handled in a single call. The API returns a sourceFileId immediately and starts the pipeline asynchronously. |
Step 2 — Pipeline splits, validates, and clears conditions The splitter separates individual documents; the classifier identifies each one; the extractor pulls signature presence, dates, and notary data; the LOS validator checks extracted values against investor rule configurations and flags VALIDATION_ERROR events for any failed checks. |
Step 3 — Validated documents push to Encompass eFolders Documents that pass validation are pushed automatically to the correct Encompass eFolder locations. Retrieve the condition clearance status via GET /loans/{loanId}/reports and confirm delivery via GET /reports/loan-processing once docPushDateTime is populated. |
What you get
| 4× post-close productivity Analysts handle exceptions, not full packages. Routine signature, date, and presence checks run automatically. The throughput improvement is consistent — not dependent on reviewer experience or workload. |
Avg 2 min 30 sec per document Average processing time from source file pickup to Encompass push across the full pipeline. For a 100-page closing package split into 20 documents, the entire file is validated and delivered in under an hour without human involvement. |
Investor-grade QC export The AI performance report at GET /reports/ai-performance provides one row per extracted field per document — classification confidence, extraction confidence, and whether manual correction was applied. Ready for investor audit delivery without additional preparation. |
Common questions
Does MOS handle investor-specific validation rules?
Yes. Validation rules are configured at the tenant level via loan configuration objects. Each configuration maps a set of required documents, signature placement requirements, notary completeness checks, and date constraints to a specific investor or product type. The LOS validator stage runs every uploaded closing package against the applicable configuration automatically. You can retrieve the active configurations and their associated reports via GET /loans/{loanId}/configurations/{configurationId}/reports.
What happens if a signature is missing?
The extractor checks signature presence on every document that requires it. A missing or incomplete signature generates a VALIDATION_ERROR event in the analytics stream and creates a review task for that document. The document will not reach EXPORTED status until the exception is resolved — either by a human reviewer correcting the extracted value via PUT /extractedData/documents/{documentId} or by uploading a corrected document. The exception list is queryable at any time via GET /loans/{loanId}/reports.
How do I get the condition clearance report?
Call GET /loans/{loanId}/reports for the full list of reports associated with the loan, then retrieve a specific report at GET /loans/{loanId}/reports/{reportId}. You can also create a new condition checklist report on demand via POST /loans/{loanId}/reports. For investor-grade QC export, use GET /reports/ai-performance — this returns per-field extraction detail across all documents, suitable for direct inclusion in investor delivery packages.
Can MOS handle packages from any title company?
Yes. MOS accepts any standard closing package PDF delivered from settlement or title agents — there is no requirement for title companies to change their workflow or format. The splitter service handles variable page counts and mixed document layouts. Packages of 100–300 pages are processed within the standard pipeline SLA. The only requirement is that the file is uploaded as a PDF; multi-file packages can be submitted in a single multipart request.
Key endpoints
Ingest closing packages, automate splitting and validation, and surface only true exceptions for review.
/loans/{loanId}/source-files
Upload closing package PDF(s)
/loans/{loanId}/documents
List all split and classified closing documents
/extractedData/documents/{documentId}
Get extracted values (signatures, dates, notary data)
/loans/{loanId}/reports
Get validation and condition clearance reports
/loans/{loanId}/reports
Create a condition checklist or audit report
/reports/ai-performance
Field-level extraction detail for investor QC export
Loan Setup & Data Quality
Turn uploads into clean, decision-ready files automatically.
Underwriting Automation
Decision-ready income and asset data without manual calculations.
Processing, Closing & Post-Close
Automate condition matching, closing review, and post-close validation.