watsonx Assistant API overview

Client applications

To build a virtual assistant or other client application that communicates with an assistant at run time, use the v2 API. You can develop a customer-facing client that can be deployed for production use, an application that brokers communication between an assistant and another service, or a testing application.

By using the v2 runtime API to communicate with your assistant, your application can take advantage of the following features:

• Automatic state management. The v2 runtime API manages each session with a user, storing and maintaining all of the context data your assistant needs for a complete conversation.

• Ease of deploying assistants. In addition to supporting custom clients, an assistant can be easily deployed to popular messaging channels such as Slack and Facebook Messenger.

• Versioning. You can save a snapshot of your content and link your assistant to that specific version. You can then continue to update your development version without affecting the production assistant.

• Search capabilities. The v2 runtime API can be used to receive responses from the search integration. When a query is submitted that your assistant can't answer, the search integration can find the best answer from the configured data sources.

For more information, see the watsonx Assistant v2 API reference.

The watsonx Assistant v1 API supports the /message method that sends user input directly to the workspace used by a dialog. The v1 runtime API is supported primarily for compatibility. If you use the v1 /message method, you must implement your own state management, and you cannot take advantage of versioning or any of the other features of an assistant.

Authoring applications

The v1 API provides methods that enable an application to create or modify dialog skills, as an alternative to building a skill graphically by using the classic Watson Assistant user interface. An authoring application uses the API to create and modify skills, intents, entities, dialog nodes, and other artifacts that make up a dialog skill. For more information, see the v1 API Reference.

The v1 authoring methods create and modify workspaces rather than skills. A workspace is a container for the dialog and training data (such as intents and entities) within a dialog skill. If you create a new workspace by using the API, it appears as a new dialog skill in the classic Watson Assistant user interface.

For a list of the available API methods, see API methods summary.