In the ever-evolving world of import customs clearance and supply chain management, efficiency is key. Manual data entry from shipping documents like House Bills of Lading (HBL), Commercial Invoices (CI), and Packing Lists (PL) is time-consuming and prone to errors. To address this, we are introducing an advanced machine learning-powered Autofill feature designed to streamline the data extraction process. By simply uploading relevant documents, users can experience seamless automation, reducing manual workload and enhancing accuracy.
Note: The Autofill feature is currently available for the commercial entities module in beta. General availability across all modules is planned by the end of Q4 2025, with an expected accuracy of over 90% for all document types.
Step 1: Open the Autofill form
Click the "Autofill" button in a new commercial entity importer form (or in a non-importer, as the Autofill feature works the same in both commercial entity forms).
In a new Importer form:
In a new Non-Importer form:
Step 2: Upload your document
Now, it's time to upload your document for analysis. Click the "Upload Document" button and select your file:
Autofill accepts file formats in .pdf, .jpeg, .jpg, .png, .bmp, .tiff, and .heif. It is currently trained only to analyze and extract data from a House Bill of Lading.
Step 3: Run the analysis
Click the "Analyze" button at the bottom right to start processing the document:
Our machine learning model first tries to classify the document. If you have uploaded a correct document type that our model currently accepts (House Bill of Lading in this case), you will see a message like the one below with the confidence level from the model:
If you submitted a different document type that our model currently does not accept, you will see an error message like the one below:
Step 4: Select the entity
Autofill tries to extract below entities that are standard in each international shipping:
Shipper
Consignee
Notify party
Forwarding agent
Destination agent
Once Autofill extracts the data, you will see the entities extracted in the below format.
Here, you can see the field name, value, and confidence level for the extracted value:
You can sort the results in ascending or descending order by column by just clicking the respective column name:
Then, to select the entity for which you are creating their profile in the commercial entities module, scroll down and click the "Use This Entity" button at the bottom left:
The form will be auto-populated with the extracted data like below:
While the Autofill feature extracts most data automatically, you may need to manually fill in any remaining fields before submitting your entry. We are continuously improving our machine learning model to recognize fields that require contextual understanding, such as entity type, address type, etc., to further minimize manual input.
Always double-check the extracted data before submission. Although our machine learning model is consistently learning and improving, AI-extracted information may still be erroneous or incorrect. Your review ensures accuracy and completeness.
The Autofill (beta) feature in pCustoms significantly enhances efficiency by automating data extraction from shipping documents, reducing manual entry, and minimizing errors. By following the steps outlined in this guide, you can quickly auto-populate a new commercial entity form using extracted data from a House Bill of Lading. While our machine learning model continues to improve, certain fields may still require manual input, especially those needing contextual understanding. To ensure accuracy, always review extracted data before submission. We appreciate your feedback as we refine this feature, with the goal of achieving over 90% accuracy by the end of Q4 2025.