How do I access Amazon Sagemaker through React Native? - amazon-web-services

Question

I am currently using react-native to build a mobile application. I need to access a machine learning model in order to send pictures for segmentation. I want to be able to receive a segmented picture back to have the background of the picture cut out. I am trying to use Amazon Sagemaker (because it seems to be a easy to work with package, but if there are other ways to do it, please let me know).
On this Sagemaker quick-start guide, on step 5a, it states:
5a. To deploy the model on a server and create an endpoint that you can access, copy the following code into the next code cell and select Run:
xgb_predictor = xgb.deploy(initial_instance_count=1,instance_type='ml.m4.xlarge')
I want to host everything on AWS and not have to run a separate server. What service/process could I use that would allow me to create an endpoint that I can access through react-native?

Answer 1

To summarize the conversation in the comments:
Once you have your model trained, tuned, and deployed (which is not a simple process), you can call the endpoint of the model using the AWS SDK for JavaScript, that you install by:
npm install aws-sdk
var AWS = require('aws-sdk/dist/aws-sdk-react-native');
you include in the HTML as:
<script src="https://sdk.amazonaws.com/js/aws-sdk-2.538.0.min.js"></script>
And when you want to call the endpoint you invoke it like that:
var params = {
Body: Buffer.from('...') || 'STRING_VALUE' /* Strings will be Base-64 encoded on your behalf */, /* required */
EndpointName: 'STRING_VALUE', /* required */
Accept: 'STRING_VALUE',
ContentType: 'STRING_VALUE',
CustomAttributes: 'STRING_VALUE'
};
sagemakerruntime.invokeEndpoint(params, function(err, data) {
if (err) console.log(err, err.stack); // an error occurred
else console.log(data); // successful response
});
You can check out the Amplify Library that can take some of the heavy liftings such as getting IAM permissions to call the API, a user log in and many others.