Documents

• A Document can be a .pdf or a .jpeg file along with some meta information.

Lucidtech delivers services that helps you control and automate the flow of your documents, and a Document is therefore an important concept, and in this introduction you will see how a Document can be created, controlled and used together with Batches, Consents, Predictions, and Models.

Creating a Document

The simplest way to create a Document is to use the CLI and use the path of the PDF or JPEG that you would like to upload.

>> las documents create path/to/my/document.pdf
{
  "documentId": "las:document:84ed1bb2d2634072bd3134274ed56ebe",
  "contentType": "application/pdf"
}

Use this documentlId along with a modelId to make a prediction on the document. See predictions for more details.

Now let's say you have several documents that you want to group together with a purpose of constructing a dataset for training a model. This is where Batches enter the picture

>> las batches create --name train --description "documents for training a new model"
{
  "batchId": "las:batch:84ed1bb2d2634072bd3134274ed56ebe",
  "name": "train",
  "description": "documents for training a new model"
}
​
#### A document can now be created with a belonging `batchId`. 
​
>> las documents create path/to/my/document.pdf --batch-id las:batch:84ed1bb2d2634072bd3134274ed56ebe 
{
  "documentId": "las:document:0e62b572139b43179076323bc35b220e",
  "contentType": "application/pdf",
  "batchId": "las:batch:84ed1bb2d2634072bd3134274ed56ebe"
}

The exact same can be done for Consents, but the purpose is to separate customers data rather that grouping them together for a training purposes. Both Batches and Consents can be used as input when listing documents;

>> las documents list --consent-id las:consent:<hex-uuid>
>> las documents list --batch-id las:batch:<hex-uuid>

Attaching Ground Truth to a document

In order to train or evaluate a model we need a ground truth along with each document. See our tutorial on data training for more details.

The ground truth of a document can be provided as additional info when we create it, or it can be appended afterwards. Either way the syntax is pretty much the same:

>> las documents create path/to/document.pdf --fields amount=100.00 due_date='2021-05-20'
>> las documents update <document-id> --fields amount=100.00 due_date='2021-05-20'

By providing this information we are able to train our models by comparing our predictions to the ground truth.