Can't select Azure AI Search service

[MarcGiannuzzi]

Hi,

While trying to create an index for my Mistral AI deployment, I am facing this issue: The dropdown below "Select Azure AI Search service" does not display any AI search service.

However I do have one. That AI search service is being used in the Open AI Azure service (oai.azure.com) and the connnection works well. Thus, wether it is a Mistral AI deployment, or a Open AI deployment, I just can't connect both of those to my search ai service index.

Please can you provide some feedback on why the search service is not displayed ?

Also, I am very interested to know how to do the same actions with some python implementation. Note: (I can communicate with the LLM, but that one does not focus on my actual data in Azure AI Studio, python)

Thank you for your response,

---

Document Details

⚠ Do not edit this section. It is required for learn.microsoft.com ➟ GitHub issue linking.

• ID: 91e98025-1c94-89a4-8192-d4c57d46578e
• Version Independent ID: 7e7a25f3-0c27-84d4-1bd2-cb647b0191b3
• Content: How to create vector indexes - Azure AI Studio
• Content Source: articles/ai-studio/how-to/index-add.md
• Service: azure-ai-studio
• GitHub Login: @eric-urban
• Microsoft Alias: eur

[TPavanBalaji]

@MarcGiannuzzi Thanks for your feedback! We will investigate and update as appropriate.

[hp1911]

I am facing the exact same issue.

Any resolution? I've created an Azure AI Search Service, it doesn't show up though.

[Naveenommi-MSFT]

@MarcGiannuzzi Thank you for bringing this to our attention. I've delegated this to content author @eric-urban, who will review it and offer their insightful opinions.

[Naveenommi-MSFT]

@eric-urban Could you please review add comments on this, update as appropriate.

[levyuk]

I'm also having this issue which previously didn't occur. Happens with a new ai project and new search service in a new resource group.

[absolute-tk]

This happens to my first ever project. From previous comments, I guess I did nothing wrong.🙂

[hp1911]

Okay I found out the solution (what worked for me), it was very simple fix by going to Project Settings -> Adding Connection of Azure AI Search. Hope it helps!

[absolute-tk]

Okay I found out the solution (what worked for me), it was very simple fix by going to Project Settings -> Adding Connection of Azure AI Search. Hope it helps!

This works!!!🤩

[MarcGiannuzzi]

Hi,

Thank you @hp1911 for your answer !

However I am still having trouble connecting my deployment to my search index:

AI Studio seems to not be able to connect to the search index data.

Is the same for you ?

[justuswolff-audit]

@MarcGiannuzzi I am facing exactly the same issue and couldn't find a way to resolve it? Did you find a solution?

[MarcGiannuzzi]

Hi @justuswolff-audit ,

Nope, still struggling to connect the search service to the deployment; the selection of the index seems impossible.

Did someone success in doing that ?

[justuswolff-audit]

I looked into it and found following issue:

It looks like the base url is not added correctly for the API call. I got the API call to run in postman with this base url: [service-name].search.windows.net/. Where would I open an issue for that @eric-urban?

[alexisfcote]

Same issue here

[caleb-collar]

Also having the same issue as @alexisfcote & @MarcGiannuzzi

I will mention that my Azure AI search is on the free tier.

[justuswolff-audit]

FYI: Started working for me just now.

[MarcGiannuzzi]

Hi @justuswolff-audit ,

I can also choose my azure search index now, thank you for your notification !

However, when sending a request, the /completions API return 500 error.

Is it the same for you ?

Also, does anyone has some example of how to use the same connection with the Python SDK ?

[justuswolff-audit]

Hi @MarcGiannuzzi, for me it is still creating the index which is kind of odd. I guess it is still not working as expected. I started using the Python API as well, that seems to work. See the code below :)

import os

from dotenv import load_dotenv
from openai import AzureOpenAI

load_dotenv()

deployment = "gpt-35-turbo-16k"

client = AzureOpenAI(
    azure_endpoint=os.getenv("AZURE_OPENAI_ENDPOINT"),
    api_key=os.getenv("AZURE_OPENAI_API_KEY"),
    api_version="2024-02-01",
)

completion = client.chat.completions.create(
    model=deployment,
    messages=[
        {
            "role": "system",
            "content": """
            You are a useful assistant. 
            """,
        },
        {
            "role": "user",
            "content": "Hi",
        },
    ],
    extra_body={
        "data_sources": [
            {
                "type": "azure_search",
                "parameters": {
                    "endpoint": os.environ["AZURE_AI_SEARCH_ENDPOINT"],
                    "index_name": os.environ["AZURE_AI_SEARCH_INDEX"],
                    "authentication": {
                        "type": "api_key",
                        "key": os.environ["AZURE_AI_SEARCH_API_KEY"],
                    },
                },
            }
        ],
    },
)

Hope this helps!

[eric-urban]

please-close