Integrations
Excel integration for financial spreading
How a SpreadSpace spread comes out as Excel: two workbooks of statements and tables built on request, CSV and JSON exports, and XLSX bank statement bundles.
Every lender keeps part of its credit work in Excel: the memo exhibit a committee expects, the covenant model a relationship manager maintains, the portfolio sheet a credit officer rolls up each quarter. A spread that cannot leave the system it was built in forces that work to be retyped. SpreadSpace treats Excel as a destination in its own right.
A loan’s spread comes out as two workbooks, one of statements and one of tables, built from the loan’s data as it stands when you ask. A document’s extracted lines come out as CSV or JSON, a set of documents bundles in the background, and bank statements bundle as XLSX. This guide says what each file holds, how it stays current and where it fits beside the API and the origination system handoffs on this site.
The two workbooks
Both workbooks are built from the loan’s data as it stands when you ask, so there is nothing to request ahead of time. The statements workbook holds every line of each statement with its derivation indented beneath it: the profit and loss statement and the balance sheet on your chart of accounts, the cash flow statement as printed, and every cash flow measure after them. The tables workbook holds every table the spread builds for each filer on the loan, followed by the same measures.
| File | What it holds | How it is built |
|---|---|---|
| Statements workbook | Every line of the profit and loss statement and the balance sheet on the chart of accounts, the cash flow statement as printed, and every cash flow measure, each with its derivation indented beneath it. | On request, from the loan as it stands; rebuilt whenever the loan’s data has changed since the last. |
| Tables workbook | Every table the spread builds, for each filer on the loan, followed by every cash flow measure. | On request, independently of the statements workbook. |
| Document export, CSV | One document’s extracted line items as flat rows. | Synchronously, up to the line-item cap; above it, a bundle. |
| Document export, JSON | The same report body the extraction read endpoints return. | Synchronously, as a file attachment. |
| Bundle | Line items from a chosen set of documents on one loan. | In the background as CSV or JSON, or XLSX for bank statements; polled, then downloaded. |
Money is written as currency in both, so a cell adds up and formats as a number instead of a string that has to be cleaned. A derived figure’s rows sit indented under it, which is the layout an analyst would build by hand for a committee exhibit.
How a workbook stays current
Ask for a workbook and SpreadSpace hands back a link, or says one is being built and when to ask again. When the loan’s data changed after the last build, the link still opens the older workbook, the answer says it is stale, and a newer one is already on the way, so a scheduled job that refreshes an exhibit never blocks and never files a workbook it was not told was behind. The two workbooks are built independently, so one being rebuilt never holds up the other.
The finalized spread’s figures also live apart from any file, in the loan’s attribute snapshot, frozen at a version number when the analyst finalizes. A model that must tie to the approved figures reads the snapshot. A model that needs the whole spread laid out reads the workbook.
Document exports and bundles
One document’s extraction exports on its own as CSV, flat line-item rows, or JSON, the same report body the extraction read endpoints return. A set of documents on one loan bundles in the background: the request answers at once with an export to poll, the export reports its progress as documents complete, and when it succeeds it carries a download link. Bank statements bundle as XLSX, every other document type as CSV or JSON.
A document export is the extraction as data. The document itself stays one click away in the workspace, where every figure opens its source page with the printed line boxed, so a reviewer who questions a cell in the sheet can see where it came from.
Name the accounts the model expects
The statements workbook prints the profit and loss statement and the balance sheet on the workspace’s chart of accounts, and the chart is the lender’s to shape: accounts are renamed, added or folded, and caption rules decide where a printed line lands. A model that expects the bank’s own account names gets them because the chart carries them, not because someone retyped the sheet.
For a model keyed on codes rather than names, the export map gives each account the code the target system uses, and the mapped statements read through the API carry those codes on every account. The chart of accounts mapping page describes both.
Statement mapping and the export map in the API documentation
Set up the handoff
Create an API key with spreads read for the workbooks and extractions read and export for the document exports and bundles. Request each workbook by its loan, store the link or the file where the model reads it, and request again on a schedule or on the finalize event, which arrives signed at your endpoint when an analyst approves a spread. Run the loop against the sandbox loan on a test key first.
Excel is one of several destinations. The same figures reach a loan origination system through the integration guides on this site, and a developer building a model service reads the snapshot directly, as the API guide describes.
API keys and scopes in the documentation