Edit

Share via

ai_chat_completion_prompt plugin (preview)

Applies to: ✅ Microsoft Fabric ✅ Azure Data Explorer

The ai_chat_completion_prompt plugin enables generating chat completions using language models, supporting AI-related scenarios such as conversational AI and interactive systems. The plugin uses the in Azure OpenAI Service chat endpoint and can be accessed using either a managed identity or the user's identity (impersonation).

The ai_chat_completion_prompt plugin enables generating chat completions using language models, supporting AI-related scenarios such as conversational AI and interactive systems. The plugin uses the in Azure OpenAI Service chat endpoint and can be accessed using the user's identity (impersonation).

Prerequisites

  • When using managed identity to access Azure OpenAI Service, configure the Managed Identity Policy to allow communication with the service.

Syntax

evaluate ai_chat_completion_prompt (Prompt, ConnectionString [, Options [, IncludeErrorMessages]])

Learn more about syntax conventions.

Parameters

Name Type Required Description
Prompt string ✔️ The prompt for generating chat completions. The value can be a column reference or a constant scalar.
ConnectionString string ✔️ The connection string for the language model in the format <ModelDeploymentUri>;<AuthenticationMethod>; replace <ModelDeploymentUri> and <AuthenticationMethod> with the AI model deployment URI and the authentication method respectively.
Options dynamic The options that control calls to the chat model endpoint. See Options.
IncludeErrorMessages bool Indicates whether to output errors in a new column in the output table. Default value: false.

Options

The following table describes the options that control the way the requests are made to the chat model endpoint.

Name Type Description
RetriesOnThrottling int Specifies the number of retry attempts when throttling occurs. Default value: 0.
GlobalTimeout timespan Specifies the maximum time to wait for a response from the AI chat model. Default value: null.
ModelParameters dynamic Parameters specific to the AI chat model. Possible values: temperature, top_p, stop, max_tokens, max_completion_tokens, presence_penalty, frequency_penalty, user, seed. Any other specified model parameters are ignored. Default value: null.
ReturnSuccessfulOnly bool Indicates whether to return only the successfully processed items. Default value: false. If the IncludeErrorMessages parameter is set to true, this option is always set to false.

Configure Callout Policy

The azure_openai callout policy enables external calls to Azure AI services.

To configure the callout policy to authorize the AI model endpoint domain:

.alter-merge cluster policy callout
```
[
  {
    "CalloutType": "azure_openai",
    "CalloutUriRegex": "https://[A-Za-z0-9-]{3,63}\.(?:openai\\.azure\\.com|cognitiveservices\\.azure\\.com|cognitive\\.microsoft\\.com|services\\.ai\\.azure\\.com)(?:/.*)?",
    "CanCall": true
  }
]
```

Configure Managed Identity

When using managed identity to access Azure OpenAI Service, you must configure the Managed Identity policy to allow the system-assigned managed identity to authenticate to Azure OpenAI Service.

To configure the managed identity:

.alter-merge cluster policy managed_identity
```
[
  {
    "ObjectId": "system",
    "AllowedUsages": "AzureAI"
  }
]
```

Returns

Returns the following new chat completion columns:

  • A column with the _chat_completion suffix that contains the chat completion values.
  • If configured to return errors, a column with the _chat_completion_error suffix, which contains error strings or is left empty if the operation is successful.

Depending on the input type, the plugin returns different results:

  • Column reference: Returns one or more records with additional columns prefixed by the reference column name. For example, if the input column is named PromptData, the output columns are named PromptData_chat_completion and, if configured to return errors, PromptData_chat_completion_error.
  • Constant scalar: Returns a single record with additional columns that are not prefixed. The column names are _chat_completion and, if configured to return errors, _chat_completion_error.

Examples

The following example generates a chat completion for the prompt Provide a summary of AI capabilities using the Azure OpenAI chat completion model.

let prompt = 'Provide a summary of AI capabilities';
let connectionString = 'https://myaccount.openai.azure.com/openai/deployments/gpt4o/chat/completions?api-version=2024-06-01;managed_identity=system';
evaluate ai_chat_completion_prompt(prompt, connectionString)

let prompt = 'Provide a summary of AI capabilities';
let connectionString = 'https://myaccount.openai.azure.com/openai/deployments/gpt4o/chat/completions?api-version=2024-06-01;impersonate';
evaluate ai_chat_completion_prompt(prompt, connectionString)

The following example sends a separate prompt for each row to the Azure OpenAI chat completion model.

let connectionString = 'https://myaccount.openai.azure.com/openai/deployments/gpt4o/chat/completion?api-version=2024-06-01;managed_identity=system';
let options = dynamic({
  "RetriesOnThrottling": 1,
  "GlobalTimeout": 2m,
  "ModelParameters": {
    "temperature: 0.7
  }
});
datatable(Prompt: string)
[
  "Provide a summary of AI capabilities",
  "What is the answer to everything?",
  "What is 42?"
]
| evaluate ai_chat_completion_prompt(prompt, connectionString, options , true)

Related content

  • Last updated on