Building an API

At this point we've created and deployed a model, and we've constructed a Lambda function that can take care of processing user data, sending it off to our deployed model and returning the result. What we need to do now is set up some way to send our user data to the Lambda function.

The way that we will do this is using a service called API Gateway. Essentially, API Gateway allows us to create an HTTP endpoint (a web address). In addition, we can set up what we want to happen when someone tries to send data to our constructed endpoint.

In our application, we want to set it up so that when data is sent to our endpoint, we trigger the Lambda function that we created earlier, making sure to send the data to our Lambda function for processing. Then, once the Lambda function has retrieved the inference results from our model, we return the results back to the original caller.