Quick extraction

After you've configured extractions for your custom documents or for out-of-the-box supported documents, you can upload documents to the Sensible app and download the extracted data as Excel files.

Note: A bulk file upload feature is coming soon.

In this tutorial you'll learn to:

  • upload documents to the Sensible app's Quick extraction tab
  • automatically extract from the documents using pre-existing document types and configs
  • download the extracted data as Excel.

Take the following steps:

  1. Navigate to the Quick Extraction tab.

  2. In the Document type dropdown in the right pane, select your document type.

    If you haven't created your own document type, select sensible_instruct_basics / Auto select . The sensible_instruct_basics document type contains interactive examples for bank statements, resumes, and contracts. If you specify Auto select, Sensible automatically detects the best-fitting extraction configuration, or "config", for the document you upload.

Click to enlarge

  1. Click Upload document and select your example document. For a list of supported file types, see Supported file types. Note that Sensible automatically OCRs documents as needed, except in advanced cases.

​ If you don't have an example document, use the following document with the sensible_instruct_basics document type :

Example documentDownload link
  1. Click Run Extraction.

  2. Sensible displays the extracted data as JSON in the right pane. Click the Download excel to convert the extracted document data to Excel:

Click to enlarge

The following spreadsheet shows output for the example bank statement. The first tab contains fields with single values, for example the start date field. Succeeding tabs contain fields with table output, for example, the accounts list table.

Note Each downloaded Excel file contains the data from one document. To combine extracted documents into one Excel file, use the Sensible API or Sensible SDKs.

Next

For more information about how Sensible converts JSON document extractions to Excel, see SenseML to spreadsheet reference.