Skip to main content

​
Getting to the upload page

In the Loan Pipeline, click Upload CSV in the page header.

​
Upload types

Before selecting a file, choose an upload type. There are three options:

​
Selecting a file

After choosing an upload type, select your CSV file by either:
  • Dragging and dropping the file onto the drop zone
  • Clicking the drop zone to open a file browser
File requirements:
  • File format: .csv only
  • Maximum file size: 10 MB
  • Must have a header row

​
Preview and column mapping

After selecting a file, Keystone parses it and shows a preview:
  1. Column mapping summary — Shows which CSV columns were recognized and mapped to loan fields. Recognized columns appear as colored pills.
  2. Ignored columns — Collapsible section showing CSV columns that were not recognized. These columns are skipped during import.
  3. Data preview table — Shows the first several rows of mapped data so you can verify the mapping is correct.
  4. Validation errors — Any parsing issues found during preview.

​
Replacing a file

If the mapping looks wrong, click Replace to select a different file, or click X to clear the selection and start over.

​
Column reference — Pipeline and Analytics uploads

The CSV importer uses flexible header matching. It recognizes many common header names from popular LOS exports (including Arive, LendingPad, Encompass, and others). Headers are matched case-insensitively.

​
Loan identifiers

​
Borrower

​
Property

State values accept both full state names (e.g., “California”) and abbreviations (e.g., “CA”). Occupancy values are mapped flexibly: “Primary Residence”, “Owner Occupied”, “Second Home”, “2nd Home”, “Investment Property” all map correctly.

​
Loan details

​
Financial

Dollar values are cleaned automatically — currency symbols ($), commas, and whitespace are stripped before parsing.

​
Timeline dates

Dates are parsed flexibly. The importer accepts US format MM/DD/YYYY (common in LOS exports) and falls back to standard date parsing.

​
Team members

​
Agent information

​
Column reference — Fee Report uploads

Fee report uploads only require the broker loan number to match existing loans. All other columns update fee fields:

​
How data is mapped

​
New loans vs updates

For Pipeline and Analytics uploads, the importer uses the Broker Loan Number as the match key:
  • If a loan with that broker loan number already exists in your company, the row updates the existing loan with any non-empty values from the CSV.
  • If no matching loan exists, a new loan is created.
For Fee Report uploads, only existing loans are updated. Rows with unrecognized broker loan numbers are skipped with an error.

​
Employee matching

Loan officers, LOAs, and processors are matched by:
  1. Email (exact match, case-insensitive) — checked first
  2. Full name (exact match, case-insensitive) — checked second
If no match is found and the CSV includes a name, Keystone auto-creates the employee with the appropriate role. The upload result summary lists any auto-created employees.

​
Lender matching

Lenders are matched by name (case-insensitive). If no match is found, a new lender is auto-created. The upload result summary lists any newly created lenders.

​
Lead source matching

Lead sources are matched by name (case-insensitive). If no match is found, a new lead source is auto-created.

​
Status derivation

If the CSV does not include an explicit status column, the importer derives the status from the timeline dates. It checks dates in reverse order of the pipeline (from Adverse down to App Intake) and sets the status to the most advanced milestone that has a date.

​
Processing the upload

After reviewing the preview:
  1. Click Upload {N} rows to start the import.
  2. The system processes all rows and displays results:
    • Success banner: Shows count of loans created and updated.
    • Errors: Lists any rows that could not be processed, with row number and reason.
    • Warnings: Lists non-fatal issues (e.g., unrecognized enum values).
    • New lenders: Lists any lenders that were auto-created.
    • New employees: Lists any employees that were auto-created.
  3. If there are no errors, you are automatically redirected to the pipeline.
  4. If there are errors, you remain on the upload page to review them.

​
Tips for preparing your CSV

  1. Always include a Broker Loan Number column. This is the only required column and is used to match existing loans.
  2. Use headers from your LOS export as-is. The importer recognizes headers from Arive, LendingPad, Encompass, and other common systems. You usually do not need to rename columns.
  3. Leave cells empty for unknown values. Empty cells are skipped; they do not overwrite existing data.
  4. Dollar amounts can include $ and commas. Values like $350,000.00 are parsed correctly.
  5. Dates can be in MM/DD/YYYY format. This is the most common format in LOS exports.
  6. Check the column mapping preview before uploading. Make sure your important columns (loan number, LO name, status) are recognized.
  7. Unrecognized columns are safely ignored. Extra columns in your CSV will not cause errors.
  8. Maximum file size is 10 MB. For very large imports, consider splitting into multiple files.