Accessing Extracted Fields from IXP Document Extraction in Maestro

Help Document Understanding
1

Hi everyone,

I am working with UiPath Maestro and using a Document Extraction service task with a previously deployed IXP model.

The extraction itself seems to work correctly
and the output is not empty. However, I am struggling to access the extracted values in subsequent workflow/service tasks.

For example, I tried expressions like:

“vars.output?.ExtractionResult?.ResultsDocument?.Fields?.[0]?.Value”

but the result is always null, even though the extraction task completed successfully and returned data.

My questions are:

  1. What is the correct way to access individual extracted fields (e.g. Vendor,
  2. Total_Amount, Date) from a Document Extraction output in Maestro?
  3. How can I pass the complete extraction result to the next task in the BPMN workflow?
  4. Is there a recommended way to store the entire extraction output in a process variable and reuse it later?

Any examples or best practices would be greatly appreciated.

Thanks!

image (1)
image (1)376×147 5.46 KB

image
image471×823 30.5 KB

2

Hi @iboti57

The issue is most likely with the output structure. In Maestro, the Document Extraction output is available through the node output, for example:

$vars..output

Also, Value is not directly under Fields. You need to access it through Values[0].Value.

If Vendor, Total_Amount, Date are inside the Expense Details field group (as shown in your screenshot), you first need to access that field group and then its Components.

I would suggest checking the actual output of the extraction node and then mapping the required values to process variables. You can then use those variables in the following tasks.

So instead of using vars.output…, try using the Output of the Document Extraction node from the expression/variable picker.

3

I checked the raw output and the extracted data is actually under the Values array, not directly under Fields.

My issue is that the Expression Picker only exposes a flattened structure and does not show Values or Components. When I try to access them manually in the JS editor, I get a type error.

image (2)
image (2)732×425 37.4 KB

image (3)
image (3)786×420 31.6 KB

4

Hi @iboti57,

I have actually implemented this in one of the projects. Here are the snaps to find how I achieved to get the details as an output.

This is my taxonomy in IXP.

image
image1246×559 42.4 KB

For example, to extract Age, I have used this syntax.

image
image981×196 11.8 KB

This is to extract Name.

image
image985×177 12 KB

The structure of the JavaScript code depends on how you actually define the taxonomy of your IXP document. If it’s a straight taxonomy similar to mine above, you can take a reference of my above snaps.

I hope this fixes your issue.

Cheers!

5

Hi @iboti57

yes, that makes sense. If the raw output contains Values but the expression picker exposes only the flattened Fields type, then the issue seems to be with the exposed schema/type rather than the actual extraction result.

I would try using the JSON editor or map the complete output to a JSON/object variable first, and then access Values[0].Value from that variable. The current Maestro docs also show that Values is part of the actual extraction response structure.

Also, make sure you’re using the current output variable name, since the recent Maestro update renamed ExtractionResult to result.