Can I add OpenAPI specification to a webservice deployed with AzureML in AKS?

Question

I'd like to deploy a machine learning service using AzureML on AKS. I also need to add some OpenAPI specification for it.

Features in https://learn.microsoft.com/en-us/azure/machine-learning/how-to-deploy-azure-kubernetes-service?tabs=python are neat, but that of having API docs/swagger for the webservice seems missing.

Having some documentation is useful especially if the model takes in input several features of different type.

To overcome this, I currently get models trained in AzureML and include them in Docker containers that use the python FastAPI library to build the API and OpenAPI/Swagger specs, and those are deployed on some host.

Can I do something equivalent to this with AKS in AzureML instead? If so, how?

Answer 1

@Davide Fiocco The deployments of Azure ML provide a swagger specification URI that can be used directly. The documentation of this is available here. You can print your swagger_uri of the web service and check if it confirms with the specifications you are creating currently.

If the above response helps, please accept the response as answer. Thanks!!