Help Centre › Building Processes
This article describes the RPA Outlook functionality to extract e-mail data using AI.
For a full list of available functions, please read OUTLOOK functions.
Description
The OUTLOOK Extract e-mail data function uses AI to extract structured data from the currently selected e-mail body. Specify the field names you want to extract (e.g. invoice number, amount, due date, vendor) and the AI will parse the e-mail content and return the values.
Setting it up
To use the OUTLOOK Extract e-mail data function, follow these instructions:
{{popups:rpas/add:action=outlook-extract-data}}
Once on a process map:
- Navigate to the process where you would like to add the OUTLOOK Extract e-mail data function
- Click on the green button's drop-down (right side of the green button labelled "Process")
- On the menu that appears, click on the "+ Activity" item
- From here you can select the OUTLOOK Extract e-mail data option
- Click "Add" once you are done
Input parameters
| Parameter | Required | Type | Default | Description |
|---|---|---|---|---|
| Fields | Yes | Table | - | List of field names to extract, e.g. invoice_number, amount, due_date, vendor |
| Instructions | No | Textarea | - | Additional extraction context or rules for the AI |
| Variable | Yes | Text | - | Variable to store the extracted data dictionary |
Output variables
| Variable | Type | Description |
|---|---|---|
| Variable | Dictionary | Extracted fields as key-value pairs (null for fields not found) |
Errors
| Error | Description |
|---|---|
| Outlook not open | Outlook is not running |
| No e-mail selected | No e-mail is currently selected |
| AI processing error | The AI could not process the e-mail content |
Tips
- Use clear, descriptive field names for better extraction accuracy
- Add instructions to guide the AI for complex or ambiguous data
- Fields not found in the e-mail will have null values
- Works best with structured e-mails like invoices, orders, and confirmations