Automated PDF Invoice Extraction to Xero (Make.com)

1. **Establish the Trigger Source**: Since PDFelement is desktop-based, configure it to save outputs to **Wondershare Document Cloud** or a linked folder (Google Drive/Dropbox). Add the 'Watch Files' module in Make.com for that specific folder. 2. **Connect Wondershare/Cloud API**: In Make.com, click 'Add', search for your cloud storage, and authenticate. Set the 'Limit' to 1 to process bundles one by one during initial setup. 3. **Leverage PDFelement AI/OCR**: Use a 'HTTP Make a Request' module or a dedicated OCR module to send the file to PDFelement’s extraction API. Ensure you request the 'Summary' or 'Markdown' output for structured text. 4. **Parse the Data**: Add a 'JSON Parser' or 'Text Parser' module. Use **Regular Expressions (Regex)** within Make.com to isolate 'Total Amount', 'Invoice Date', and 'Vendor Name' from the AI-generated summary. 5. **Establish Xero Connection**: Add a Xero module ('Create an Invoice'). Click 'Create a connection' and log in via OAuth2. Select the specific Xero Organization. 6. **Check for Existing Contact**: Add a Xero 'Search Contacts' module. Map the 'Vendor Name' from the PDF. This prevents duplicate contact creation. 7. **Implement a Router**: Use a Router to branch the logic. Path A (Contact found): Proceed to Create Bill. Path B (Contact not found): Create Contact FIRST, then Create Bill. 8. **Map Bill Details**: In the 'Create a Purchase/Bill' module, map the extracted date using `formatDate(parsed_date; YYYY-MM-DD)` and the total amount using the `parseNumber()` function to ensure Xero accepts the decimal format. 9. **Attach Original File**: Add the Xero 'Upload a File' module. Source the 'File Content' from the initial trigger module to ensure the PDF is linked to the Bill in Xero. 10. **Add Error Handling**: Right-click the Xero module and select 'Add error handler'. Choose 'Break' to store the execution for manual review if the API fails due to validation errors (e.g., missing account codes).

| Data Source (PDFelement/AI) | Xero Field | Transformation / Make.com Function | | :--- | :--- | :--- | | Extracted Vendor Name | Contact Name | `trim(vendor_name)` | | Extracted Date | Date | `formatDate(date; "YYYY-MM-DD")` | | Total Amount | Line Amount | `parseNumber(total; ".")` | | Invoice/Ref Number | Reference | `ifEmpty(ref; "No-Ref-Found")` | | AI Summary Text | Description | `substring(summary; 0; 255)` | | PDF File (Binary) | Attachment | `Upload via File Content` |

* **Data Type Mismatches**: Xero requires specific date formats (YYYY-MM-DD). Use the `formatDate` function in Make.com; otherwise, the scenario will fail with a 400 error. * **Rate Limits**: Xero has a limit of 5000 API calls per day. If processing bulk historical PDFs, use the 'Sleep' module or 'Search' filters to avoid hitting limits. * **Bundle Overhead**: Make.com charges per operation. Use a 'Filter' after the trigger to ensure only PDF files (MIME type application/pdf) proceed through the scenario to save costs. * **Auth Token Refresh**: Ensure your Xero connection is periodically checked; while Make handles refresh tokens, a manual disconnect in Xero settings will break the scenario.

- [ ] **Run Once Test**: Use the 'Run once' button and upload a real PDF to the cloud folder. - [ ] **Check Mapping**: Verify in the Make.com execution log that `parseNumber` correctly converted '$1,200.00' to '1200.00'. - [ ] **Search Logic**: Verify that if a vendor already exists, the scenario fetches the ID rather than creating a duplicate. - [ ] **File Presence**: Log into Xero and ensure the 'Files' icon on the bill shows the original PDF attachment. - [ ] **Incomplete Data Check**: Test a 'garbage' PDF to ensure your Error Handler (e.g., Slack notification) triggers correctly.