When you finish building, teaching and testing your Duet model, make it available to your client application by deploying it to the endpoint. In Duet, a deployment is a single model or set of models that are made available via a REST API endpoint. The REST endpoint takes a single text document as input and produces a JSON response classifying the text into the proper category as output. The API takes a document of 300KB as maximum size. Duet models are fully hosted and there is no on prem option currently offered, but it is on the product roadmap. Traffic received at the endpoint is stored in Duet so that you can come back and improve the model based on the real traffic that get appended every 12 hours to the unlabeled dataset associated with the model. Duet Machine Learning Operations (ML Ops) enables users to perform effective CI/CD for their models. If you update your model and publish a new version, you can roll out the new version to replace the older version on the same endpoint URL for continuous integration. Alternatively, you can publish the new version to a new endpoint so that you keep the new version to your staging environment or for A/B testing with the version of the model that is in production.
You can access an instructional video on deployments here.