Using cognitive services as a container
Using cognitive services hosted in Azure enables application developers to focus on the infrastructure for their own code while benefiting from scalable services that are managed by Microsoft. However, in many scenarios, organizations require more control over their service infrastructure and the data that is passed between services.
Many of the cognitive services APIs can be packaged and deployed in a container, enabling organizations to host cognitive services in their own infrastructure; for example in local Docker servers, Azure Container Instances, or Azure Kubernetes Services clusters. Containerized cognitive services need to communicate with an Azure-based cognitive services account to support billing; but application data is not passed to the back-end service, and organizations have greater control over the deployment configuration of their containers, enabling custom solutions for authentication, scalability, and other considerations.
Comments
Post a Comment