js.md

JavaScript

Use browser version

npm install --save las-sdk-browser

import { AuthorizationCodeCredentials } from '@lucidtech/las-sdk-browser';
import { Client } from '@lucidtech/las-sdk-core';

const credentials = new AuthorizationCodeCredentials('<apiEndpoint>', '<apiKey>', '<clientId>',  '<redirectUri>', '<launchUriFn>', '<authEndpoint>');
const client = new Client(credentials);

Use node version

npm install --save las-sdk-node

import { AuthorizationCodeCredentials } from '@lucidtech/las-sdk-node';
import { Client } from '@lucidtech/las-sdk-core';

const credentials = new ClientCredentials('<apiEndpoint>', '<apiKey>', '<clientId>',  '<clientSecret>', '<authEndpoint>');
const client = new Client(credentials);

Make a prediction on a document

Suppose we wish to run inference on a document using Lucidtech’s invoice model.

const { documentId } = await client.createDocument('<document content>', '<document mime type>', { consentId: '<consent id>' });
const { predictions } = await client.createPrediction(documentId, '<model name>');
console.log(predictions);

Set ground truth of document

When uploading data that will be used for training and evaluation, we need to provide a ground truth.

const groundTruth = [
  { 'label': 'total_amount', 'value': '240.01' },
  { 'label': 'due_date', 'value': '2020-01-31' },
];
const { documentId } = await client.createDocument('<document content>', '<document mime type>', { groundTruth });

Update an existing document

If for instance a prediction reveals incorrect values in the ground truth of a document, we can update the existing document with new ground truth values.

const groundTruth = [
  { 'label': 'total_amount', 'value': '240.01' },
  { 'label': 'due_date', 'value': '2020-01-31' },
];
client.updateDocument('las:document:<hex-uuid>', groundTruth);

{% hint style="info" %} Providing ground truth is a necessary to re-train a model whether the model got it right or wrong. So always provide the ground truth if it is available. {% endhint %}

Create a document with consent id

{% hint style="info" %} Consent ID is an identifier you can assign to documents to keep track of document ownership for your customers. {% endhint %}

const { documentId } = await client.createDocument('<document content>', '<document mime type>', { consentId: '<consent id>' });

Create a batch and associate a few documents with it

Creating a batch is a way to group documents. This is useful for specifying batches of documents to use in improving the model later.

const { batchId } = await client.createBatch(batchDescription);
await client.createDocument('<document content>', '<document mime type>', { consentId: '<consent id>',  batchId })
await client.createDocument('<another content>', '<another mime type>', { consentId: '<consent id>',  batchId })