Community

AI Configuration

Written by Tyler Shuck | Sep 23, 2025 6:41:41 PM

Introduction

This document gives a detailed view of AI configuration on Akumina and how to get configuration details. Once the customer has all the required details, can go into App manager and complete AI configuration setup.

 

"Max" Writing Assistant
The new Writing Assistant brings generative A.I. into the Voice editorial experience. Using your organization's Azure AI license, the Max Assistant will help content authors create content, apply modifiers, and use editing tools to make sure all content meets your organizational standards.

Max can also help you revise existing content in the Revise tab, which will appear when you access the Writing Assistant while there is already text in the field:

 

6.2 AI Configuration Services

Below are the services offered as part of Akumina AI services.

  • Azure Open AI Text
  • Azure Open AI Image

To enable these AI services, get each service key values as mentioned on below json format.

 

 

{

    "KernelMemory": {

      "Services": {

        "Bing": {

          "Key": "**************************************",

          "Endpoint": "https://api.bing.microsoft.com/v7.0/search?"

        },

        "AzureSpeech": {

          "Region": "eastus2",

          "Key": "**************************************"

        },

        "ContentSafety": {

          "Enabled": false,

          "ViolationThreshold": 4,

          "Endpoint": "https://*****.cognitiveservices.azure.com/",

          "Key": "**************************************"

        },

        "AzureDocumentIntelligence": {

          "Enabled": true,

          "Extensions": [ "png", "jpeg", "jpg", "xlsx" ],

          "Key": "**************************************",

          "Endpoint": "https://*******.cognitiveservices.azure.com/"

        },

        "AzureOpenAIText": {

          "Auth": "ApiKey",

          "APIKey": "**************************************",

          "Endpoint": "https://*****.openai.azure.com/",

          "Deployment": "*****",

          "APIType": "ChatCompletion",

          "MaxRetries": 3

        },

        "AzureOpenAIImage": {

          "Auth": "ApiKey",

          "APIKey": "**************************************",

          "Endpoint": "https://*****.openai.azure.com/",

          "Deployment": "*****",

          "MaxRetries": 1

        },

        "AzureOpenAIEmbedding": {

          "Auth": "ApiKey",

          "APIKey": "**************************************",

          "Endpoint": "https://*****.openai.azure.com/",

          "Deployment": "************************",

          "MaxRetries": 1

        },

        "AzureOpenAIWhisper": {

          "Auth": null,

          "APIKey": "**************************************",

          "Endpoint": "https://******.openai.azure.com/",

          "Deployment": "********",

          "MaxRetries": 3

        }

      }

    }

}

 

Azure Open AI Text

Azure OpenAI Service brings the powerful language models developed by OpenAI, such as GPT-3, Codex, and others, to the Azure platform. These models can be used for a variety of natural language processing tasks.

Steps to enable Azure Open AI Text Service.

  • Create a new Azure Open AI Service on Azure portal.

     

  • After Azure Open AI Service creation, get key and endpoint from “Keys and Endpoint” menu from left side panel.

     

  • When click on Azure AI Foundry Portal, it takes to Azure AI Foundry and there you will have provision to create new custom models and deployments.

     

     

  • Use above highlighted values on Azure Open AI Text Service json.

       "AzureOpenAIText": {

          "Auth": "ApiKey",

          "APIKey": "**************************************",

          "Endpoint": "https://*****.openai.azure.com/",

          "Deployment": "*****",

          "APIType": "ChatCompletion",

          "MaxRetries": 3

        }

Azure Open AI Image

Azure OpenAI Service, while primarily focused on natural language processing and generation tasks using models like GPT-3 and Codex, doesn't directly provide capabilities for image processing.

Steps to enable Azure Open AI Image Service.

  • Create a new Azure Open AI Service on Azure portal if required (because same Azure OpenAI can be used for Text, Image, Embedding and Whisper service).

     

  • After Azure Open AI Service creation, get key and endpoint from “Keys and Endpoint” menu from left side panel.

     

  • When click on Azure AI Foundry Portal, it takes to Azure AI Foundry and there you will have provision to create new custom models and deployments.

     

     

  • Use above highlighted values on Azure Open AI Image Service json.

       "AzureOpenAIImage": {

          "Auth": "ApiKey",

          "APIKey": "**************************************",

          "Endpoint": "https://*****.openai.azure.com/",

          "Deployment": "*****",

          "MaxRetries": 1

        }

Miscellaneous:

Other AI Configuration Services

Below are the other services offered as part of Akumina AI services.

  • Azure Speech
  • Content Safety
  • Azure Document Intelligence
  • Azure Open AI Embedding
  • Azure Open AI Whisper

Azure Speech Service

Azure Speech Service, part of Azure Cognitive Services, provides various functionalities for integrating speech capabilities into applications. It offers a range of features for speech-to-text, text-to-speech, and speech translation, among others.

Steps to enable Azure Speech Service.

  • Create a new Speech Service on Azure portal.

     

  • After Speech Service creation, get key and region from “Keys and Endpoint” menu from left side panel.

     

  • Use above highlighted values on Azure Speech Service json.

       "AzureSpeech": {

            "Region": "eastus2",

            "Key": "**************************************"

            }

Content Safety Service

Azure Content Safety Service is a set of AI-driven tools offered by Microsoft as part of Azure Cognitive Services, aimed at identifying and moderating harmful or inappropriate content in text, images, and other media. This service helps ensure that user-generated content on platforms and applications is safe and compliant with community standards and guidelines.

Steps to enable Azure Content Safety Service.

  • Create a new Content Safety Service on Azure portal.

     

  • After Content Safety Service creation, get key and endpoint from “Keys and Endpoint” menu from left side panel.

     

  • Use above highlighted values on Content Safety Service json.

       "ContentSafety": {

          "Enabled": false,

          "ViolationThreshold": 4,

          "Endpoint": "https://*****.cognitiveservices.azure.com/",

          "Key": "**************************************"

        }

Azure Document Intelligence

Azure Document Intelligence is a powerful AI service within Azure Cognitive Services that helps extract, analyze, and process information from documents. This service is designed to automate the extraction of data from various document types, allowing for streamlined data entry and improved document processing workflows.

Steps to enable Azure Document Intelligence Service.

  • Create a new Document Intelligence Service on Azure portal.

     

  • After Azure Document Intelligence Service creation, get key and endpoint from “Keys and Endpoint” menu from left side panel.

     

  • Use above highlighted values on Azure Document Intelligence Service json.

       "AzureDocumentIntelligence": {

          "Enabled": true,

          "Extensions": [ "png", "jpeg", "jpg", "xlsx" ],

          "Key": "**************************************",

          "Endpoint": "https://*******.cognitiveservices.azure.com/"

        }

Azure Open AI Embedding

Azure OpenAI Service, which integrates OpenAI's powerful models into the Azure platform, includes support for embeddings. Embeddings are a way to represent text data as vectors (numerical representations) that capture the semantic meaning of the text. These vectors can then be used for various natural language processing tasks such as similarity search, clustering, and classification.

Steps to enable Azure Open AI Embedding Service.

  • Create a new Azure Open AI Service on Azure portal if required (because same Azure OpenAI can be used for Text, Image, Embedding and Whisper service).

     

  • After Azure Open AI Service creation, get key and endpoint from “Keys and Endpoint” menu from left side panel.

     

  • When click on Azure AI Foundry Portal, it takes to Azure AI Foundry and there you will have provision to create new custom models and deployments.

     

     

  • Use above highlighted values on Azure Open AI Embedding Service json.

       "AzureOpenAIEmbedding": {

          "Auth": "ApiKey",

          "APIKey": "**************************************",

          "Endpoint": "https://*****.openai.azure.com/",

          "Deployment": "************************",

          "MaxRetries": 1

        }

Azure Open AI Whisper

Azure OpenAI Service now integrates with Whisper, OpenAI's automatic speech recognition (ASR) system. Whisper is designed to convert spoken language into written text, providing high-quality transcriptions across many languages and accents.

Steps to enable Azure Open AI Whisper Service.

  • Create a new Azure Open AI Service on Azure portal if required (because same Azure OpenAI can be used for Text, Image, Embedding and Whisper service).

     

  • After Azure Open AI Service creation, get key and endpoint from “Keys and Endpoint” menu from left side panel.

     

  • When click on Azure AI Foundry Portal, it takes to Azure AI Foundry and there you will have provision to create new custom models and deployments.

     

     

  • Use above highlighted values on Azure Open AI Whisper Service json.

       "AzureOpenAIWhisper": {

          "Auth": null,

          "APIKey": "**************************************",

          "Endpoint": "https://******.openai.azure.com/",

          "Deployment": "********",

          "MaxRetries": 3

        }