Integrating Custom Data into Azure OpenAI Chatbot.

@navba-MSFT Yes I tried this but it is not working for all the questions when I asked some specific questions regarding cricket and politics it wasn't able to give me an answer and also when I asked about the tableau calculated filed still received the same answer "We currently do not have any information".

@navba-MSFT I have tried this but it is not working for all the questions, when I have asked the question about the prime minister of India it gives me a correct answer but when I am trying to ask questions regarding the World cup or any other question like tableau it wasn't able to give me any answer all it came with " Sorry I can't have this information".

@Anjali Kumawat Thanks for your reply. Please refer to the below action plan:

Plan 1:
Please update / append your System message with the below:

Please answer using retrieved documents and combine with your internal knowledge to answer the question. Please feel free to add facts on top of information extracted from retrieved documents. Please do not forget to cite the retrieved documents for information you extract from documents. Think step by step about which information needs to be added or is missing from retrieved documents

Then click on Apply Changes as shown below:

Plan 2:

Try deploying it to your existing webapp and then test the behavior from there to isolate any issues with the chat playground.

Plan 3:

If the above plan doesn't help, I believe the behavior is related to your chat playground and not the API itself. Please try to invoke the Azure Open AI chat completion REST API as shown here from postman and check the behavior:

https://video2.skills-academy.com/en-us/azure/ai-services/openai/reference#azure-ai-search

If you want more details about the headers used in the API refer this thread.

Please remember to set the "inScope": false in your request.

Hope this helps.

@navba-MSFT I have tried all of these but it is not working.

@Anjali Kumawat Thanks for getting back. Could you please share the screenshot of the response received from the postman for Plan 3 ? Awaiting your reply.

@navba-MSFT I tried this code in python:
and it returns this : PS C:\Users\ANJALI> & C:/Users/ANJALI/AppData/Local/Programs/Python/Python312/python.exe "c:/Users/ANJALI/Documents/Trained_model_forcast_python/import requests.py"``Error: {"error":{"code":"NoCapacity","message":"The deployment does not have any capacity at the moment, please check the capacity setting and try again later."}}

and in this it comes as shown in the picture.

import requests
import json

# Set the Azure OpenAI endpoint URL
endpoint_url = ""

# Set your Azure OpenAI API key
api_key = ""

# Prepare the request headers
headers = {
    "Content-Type": "application/json",
    "api-key": api_key
}

# Prepare the request payload with a user message
payload = {
    "temperature": 0.7,
    "max_tokens": 800,
    "top_p": 0.95,
    "frequency_penalty": 0,
    "presence_penalty": 0,
    "stop": None,
    "messages": [
        {
            "role": "user",
            "content": "What are the differences between Azure Machine Learning and Azure AI services?"
        }
    ],
    "inScope": False  # Set "inScope" to false as instructed
}

# Send the POST request to Azure OpenAI API
response = requests.post(endpoint_url, headers=headers, json=payload)

# Check if the request was successful (status code 200)
if response.status_code == 200:
    # Parse and print the response
    response_data = response.json()
    print("Response:", response_data)
else:
    print("Error:", response.text)