So now that we've created a pipeline and verified our WEX, we can publish it as a REST service. So we can do that by invoking publish on our pipeline object, specify the name, Diabetes_Training_Pipeline, provide a description, and then version information. We can get hold of the rest_endpoint and have a look at the details of the rest_endpoint. To use the endpoint, client applications will need to make a REST call of HTTP. So this request must be authenticated.

So an authorization header is required or our application would require service principal with which to be authenticated. But to test this out, we'll use the authorization header from our current connection to our zero workspace which we can get use in the following code. So that's how we get that information. So this puts us in a position where we can call the REST interface.

The pipeline run asynchronously. So we would get an identifier back, which we can use to track the pipeline experiment as it runs. So that's that. Right, so for our request posts, we will pass in a rest_endpoint, the authentication header information, as well as the experiment_name. And since we have the run ID, we can use the run details we get to view the experiment as it runs.

Now, the pipeline should complete quickly and that's because in a prior step we specified that each step was to be able to use our prior reuse. We set it up so we use prior reuse. So this was done primarily for convenience at the same time in this example but in reality, you would likely want the first step to run every time in case the date has changed and trigger subsequent steps only if the output from the first step changes. So this is how we use the run ID information to get the details of the run as it runs.