Skip to content

first-profile.md

Latest commit

 

History

History
153 lines (107 loc) · 4.54 KB

first-profile.md

File metadata and controls

153 lines (107 loc) · 4.54 KB
description Create a profile to configure how AI requests are made and use it in your code.

Your First Profile

A profile combines a connection with model settings for a specific use case. Profiles let you configure temperature, model selection, and system prompts once and reuse them throughout your application.

Prerequisites

  • At least one connection created

Create a Profile

  1. In the Umbraco backoffice, navigate to the AI section
  2. Click Profiles in the tree
  3. Click Create Profile

Configure the Profile

Fill in the profile details:

Field Description Example
Name A display name for this profile "Content Assistant"
Alias A unique identifier for lookups "content-assistant"
Capability The type of AI capability "Chat"
Connection Which connection to use "OpenAI Production"
Model The specific model to use "gpt-4o"

Chat Settings

For chat profiles, you can configure:

Setting Description Default
Temperature Controls randomness (0-2). Lower is more focused. Provider default
Max Tokens Maximum tokens in the response Provider default
System Prompt Instructions sent with every request None

{% hint style="info" %} Leave settings empty to use the model's default values. Only set values you want to override. {% endhint %}

Set as Default

To use this profile when no profile is specified in code:

  1. Navigate to the AI section > Settings
  2. Select "Content Assistant" from the Default Chat Profile dropdown
  3. Click Save

{% hint style="info" %} See Managing Settings for more details on configuring defaults. {% endhint %}

Use the Profile in Code

Using the Default Profile

{% code title="ContentController.cs" %}

using Microsoft.Extensions.AI;
using Umbraco.AI.Core.Chat;

public class ContentController : UmbracoApiController
{
    private readonly IAIChatService _chatService;

    public ContentController(IAIChatService chatService)
    {
        _chatService = chatService;
    }

    [HttpPost]
    public async Task<IActionResult> GenerateSummary([FromBody] string content)
    {
        var messages = new List<ChatMessage>
        {
            new(ChatRole.User, $"Summarize this content:\n\n{content}")
        };

        // Uses the default profile configured in Settings
        var response = await _chatService.GetChatResponseAsync(messages);

        return Ok(response.Message.Text);
    }
}

{% endcode %}

Using a Specific Profile

To use a specific profile, pass the profile ID:

{% code title="ContentController.cs" %}

public async Task<IActionResult> GenerateSummary(
    [FromBody] string content,
    [FromServices] IAIProfileService profileService)
{
    var profile = await profileService.GetProfileByAliasAsync("content-assistant");

    var messages = new List<ChatMessage>
    {
        new(ChatRole.User, $"Summarize this content:\n\n{content}")
    };

    var response = await _chatService.GetChatResponseAsync(profile!.Id, messages);

    return Ok(response.Message.Text);
}

{% endcode %}

Understanding Profile Resolution

When you make a chat request:

  1. The profile is resolved by ID or alias
  2. The associated connection provides credentials
  3. Profile settings (model, temperature, and so on) configure the request
  4. Any registered middleware is applied
  5. The request is sent to the AI provider

Multiple Profiles

Create different profiles for different use cases:

Profile Use Case Model Temperature
content-assistant Content suggestions gpt-4o 0.7
code-helper Code generation gpt-4o 0.2
translator Translation gpt-4o-mini 0.3

Next Steps

Learn more about using the chat API:

{% content-ref url="../using-the-api/chat/README.md" %} Chat {% endcontent-ref %}

Or explore the core concepts:

{% content-ref url="../concepts/README.md" %} Core Concepts {% endcontent-ref %}