Try asynchronous extraction from a Sensible URL

Use Sensible’s asynchronous endpoints in production scenarios. You have two options for asynchronous processing:

• Provide your own URLs for your documents.
• Use URLs provided by Sensible for your documents.

This topic covers using URLs provided by Sensible. This is a good option if you can’t create either publicly accessible or pre-signed URLs for your documents.

For either option, you can get the results as soon as they’re ready by specifying a webhook.

Extract from a URL Sensible provides

Prerequisites

See prerequisites.

Generate the upload URL

Generate a temporary, one-time Sensible URL for a document:

1. Copy the following code sample and replace *YOUR_API_KEY* with your API key:

curl --request POST 'https://api.sensible.so/v0/generate_upload_url/tax_forms' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer YOUR_API_KEY' \
--data '{"content_type":"application/pdf"}'

2. In your Postman workspace, click Import, select Raw text, paste the code sample, and follow the prompts to import the code sample.

3. Click Send. The response looks something like the following:

JSON

  {
    "id": "14d82783-c12b-4e70-b0ae-ca1ce35a9836",
    "created": "2021-06-16T16:22:56.576Z",
    "status": "WAITING",
    "type": "tax_forms",
    "upload_url": "https://sensible-so-utility-bucket-prod-us-west-2.s3.us-west-2.amazonaws.com/EXTRACTION_UPLOAD/sensible/fc3484c5-3f35-4129-bb29-0ad1291ee9f8/EXTRACTION/14d82783-c12b-4e70-b0ae-ca1ce35a9836.pdf?AWSAccessKeyId=REDACTED&Expires=1623861476&Signature=REDACTED&x-amz-security-token=REDACTED"
}

Extract the data

Use the one-time URL you generated in the previous step to extract data from the document:

1. Copy the following code sample. Replace YOUR_UPLOAD_URL with the upload_url that you received as a response in the previous steps:

curl --request PUT 'YOUR_UPLOAD_URL' \
--data-binary '@/PATH_TO_DOWNLOADED_DOCUMENT.pdf' \
--header 'Content-Type: application/pdf' \

2. In your Postman workspace, click Import, select Raw text, paste the code sample, and follow the prompts to import to code sample.

Troubleshoot: Postman can automatically add authorization and content-type headers that cause errors:

• If Postman automatically specifies authorization for the request, then specify No Auth in the request’s Auth tab :

• If Postman automatically specifies a content type header, then ensure that the value of the Content-Type header matches that of the content_type body parameter in the request in step 1. In this case, it must be application/pdf.

3. If you haven’t already, download the following example document, which works with the example tax_forms document type you created in the Prerequisites section:

Example document	Download link

4. Correct the path to the downloaded document in your request: click the Body tab, select binary, then click Select file and select the document:

5. Click Send to send the request. The response is 200:

Retrieve extraction

To retrieve the document extraction, you have two options:

• Use the /documents endpoint. See the following steps.
• Use a webhook. See Try a webhook.

To retrieve the extraction results with the /documents endpoint, take the following steps:

1. In a previous step on this page, you generated a URL and got back a response that included an extraction ID:

JSON

  {  
    "id": "14d82783-c12b-4e70-b0ae-ca1ce35a9836"  
}  

2. Copy the document extraction id from that response.
3. Copy the following code sample and replace *YOUR_EXTRACTION_ID* and *YOUR_API_KEY*:

curl --request GET 'https://api.sensible.so/v0/documents/YOUR_EXTRACTION_ID' \
--header 'Authorization: Bearer YOUR_API_KEY'

4. In your Postman workspace, click Import, select Raw text, paste the code sample, and follow the prompts to import to code sample.

5. Click Send. The response includes a parsed_document object that looks something like the following:

JSON

  {
	"parsed_document": {
		"year": {
			"type": "string",
			"value": "2021"
		},
		"filing_status.single": {
			"type": "boolean",
			"value": true
		},
		"filing_status.married_filing_jointly": {
			"type": "boolean",
			"value": false
		},
		"filing_status.married_filing_separately": {
			"type": "boolean",
			"value": false
		},
		"filing_status.head_of_household": {
			"type": "boolean",
			"value": false
		},
		"filing_status.qualifying_widow": {
			"type": "boolean",
			"value": false
		},
		"name": {
			"type": "string",
			"value": "Connor Roy"
		},
		"ssn": {
			"type": "string",
			"value": "337-18-2333"
		}
	}
}

Assistant

Responses are generated using AI and may contain mistakes.