Gen AI

This recipe provisions Landing Zone, Data Hub, Azure Search, Open AI and Web Application integration.

You can choose whether to deploy through the console directly or on download as Terraform or GitHub to deploy later.

Estimated deployment time - 30mins

       


Components

Applying Generative AI one-click recipe will deploy following components inside of customer-provided Subscription.

GenAI image

  • Data Hub = Cosmos DB

    • NoSQL
    • Table

  • Cognitive Services

    • Open AI Service
    • Speech
    • Text Analytics

  • Azure Search (Vector Search)

  • Azure_WebApp
See also:Architecture of One-Click Recipes and Landing Zones.


Post-deployment Steps

cloudreporting image

Steps to be completed by the customer and/or Accenture team after deployment:

1.Create pipelines to save call recordings to Blob storage.

2. Create Service Bus Queues for Call Recordings and Call Transcriptions.

3. Create Azure Function Apps (or Logic Apps) to process call recordings and fetch call transcriptions making corresponding calls to Cognitive Services.

4. Connect Organizational Documents to Azure Search to create search indexes on top of internal knowledge.

5. Create scheduled Azure Function (or Logic App) which will use Open AI Service for summarization and tagging using Azure Search to provide additional context. Results will be saved to Blob storage.

6. Create Web Application and Power BI reports on top of summarizations.



Reference Architecture

GenAI One-click Recipe represents combination of Landing Zone basic services, and Core capabilities - Enterprise Data Lake and Data Hub.

CloudReporting image

Generative AI (Gen AI) one-click scenario is based on AI data Foundation. Feature Store and Unstructured engineering accelerators allow to lay the foundation of GenAI capability ready for channel consumption. The recipe allows to accelerate GenAI use cases.

Gen AI Components

1. Landing Zone: Following are the pre-requisites of having this one click recipe i.e., Having below services being provisioned New Vnet ,Keyvault, service principle, Subnet, Private end points, New Resource group and Data Persona’s.

2. In Data Hub, we take the copy of the Raw container into Cognitive services, that stores data as Archive data.

3.Cognitive Services: Cognitive Services brings AI within reach of every developer and data scientist. With leading models, a variety of use cases can be unlocked. All it takes is an API call to embed the ability to see, hear, speak, search, understand, and accelerate advanced decision-making into your apps. Enable developers and data scientists of all skill levels to easily add AI capabilities to their apps.

4. Open AI: Helps in, generative AI models with deep understandings of language and code to enable new reasoning and comprehension capabilities for building cutting-edge applications. We can detect and mitigate harmful use with built-in responsible AI and access enterprise-grade Azure security

5. Web Application: VM with web application to interact with customized open AI models.

6. Power BI is used. Create a data-driven culture with business intelligence for all at Client organization to make confident decisions using up-to-the-minute analytics.