Azure AI Discovery client library for .NET - version 1.0.0-beta.1

Azure AI Discovery lets you build investigation and knowledge-base experiences on top of your Discovery workspace. The client library exposes two top-level clients:

  • WorkspaceClient — conversations, investigations, tasks, and tools.
  • BookshelfClient — knowledge bases (create, list, index, search, delete).

Getting started

Install the package

Install the client library for .NET with NuGet:

dotnet add package Azure.AI.Discovery --prerelease

Prerequisites

Authenticate the client

The library uses Azure.Identity for authentication. DefaultAzureCredential works for most scenarios, including local development.

Uri workspaceEndpoint = new Uri("<workspace-endpoint>");
WorkspaceClient workspaceClient = new WorkspaceClient(workspaceEndpoint, new DefaultAzureCredential());

Uri bookshelfEndpoint = new Uri("<bookshelf-endpoint>");
BookshelfClient bookshelfClient = new BookshelfClient(bookshelfEndpoint, new DefaultAzureCredential());

Key concepts

  • WorkspaceClient is the entry point for workspace operations. Use its Get*Client() methods to obtain the sub-clients: GetDiscoveryConversationsClient(), GetDiscoveryInvestigationsClient(), GetDiscoveryTasksClient(), and GetDiscoveryToolsClient().
  • BookshelfClient is the entry point for knowledge-base operations via GetKnowledgeBasesClient().
  • Long-running operations (for example KnowledgeBases.CreateOrUpdate and KnowledgeBases.Delete) return an Operation; pass WaitUntil.Completed to wait for the operation to finish, or WaitUntil.Started to poll manually.

Examples

See the samples for end-to-end usage. A minimal example that creates and reads a conversation:

DiscoveryConversationsClient conversationsClient = workspaceClient.GetDiscoveryConversationsClient();

DiscoveryConversation created = await conversationsClient.CreateAsync(
    projectName: "my-project",
    investigationName: "/projects/my-project/investigations/my-investigation",
    displayName: "Getting started conversation");

DiscoveryConversation conversation = await conversationsClient.GetAsync(created.Name);
Console.WriteLine($"Conversation: {conversation.Name}");

Troubleshooting

Service operations throw a RequestFailedException on failure, with a Status code and an error Message that can be used to diagnose the failure.

Next steps

Browse the samples for more scenarios covering investigations, tasks, tools, and knowledge bases.

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.microsoft.com.

When you submit a pull request, a CLA-bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (for example, label, comment). Follow the instructions provided by the bot. You'll only need to do this action once across all repositories using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information, see the Code of Conduct FAQ or contact opencode@microsoft.com with any other questions or comments.

  • Last updated on