AI-102 EXAM QUESTIONS 2024 MICROSOFT AZURE AI 102 CERTIFICATION COURSE PART-5

This video is part five of an AI 10002 question series, focusing on a long case study question similar to those found in exams. Understanding these case studies can earn significant marks and save time. The video covers question number 41, which involves a detailed case study about a company called Wide Word Imports developing an e-commerce system. The system uses microservices and a serverless environment on Azure, targeting customers who speak English, Spanish, and Portuguese. The existing environment includes an app service plan for product management, used by employees to create and manage products. Employee inputs are in English, but product information needs to be displayed to customers in all three languages. Another app handles inventory tracking, managing order dispatch and refunds. Azure resources utilized include Azure AD for internal authentication, with all employees in the "all users" group and senior managers in the "leadership team." Azure functions are used to post to Azure Event Grid when product stock levels change, triggering actions for low or out-of-stock items. The function app uses Azure Cosmos DB with a Core SQL API, storing data for both product management and inventory tracking apps. Azure Blob storage contains global containers for product assets like images, videos, and PDFs. Cognitive Services (WWC) and Azure Video Analyzer (WWIVi) are also used. The company's requirements include leveraging AI to differentiate itself from competitors. Planned changes involve a product creation project to help employees create accessible and multilingual product entries, accelerating the process. A smart e-commerce project aims to implement cognitive search for customer browsing, and a "shopping on the go" project involves building a chatbot integrated with smart speakers. Business requirements emphasize providing a multilingual customer experience in English, Spanish, and Portuguese, scaling based on transaction volume for consistent performance, and minimizing costs. Governance and security requirements dictate that all data, processing, storage, and data centers must be located in the United States. Azure Cognitive Services must be inaccessible directly from the internet, requiring a private network. Accessibility requirements state that all images must have relevant alt text for users who have difficulty seeing or for low internet connectivity. All videos must have transcripts associated with them and included in product descriptions. All product descriptions, transcripts, and alt text must be available in English, Spanish, and Portuguese. Product creation requirements focus on minimizing the time employees spend creating products and adding assets, and eliminating the need for manual translation, opting for automatic translation of product names and descriptions into all three languages. Smart e-commerce requirements demand that cognitive search meets a 99.9% SLA for availability of searches and index writes. Users must be able to search within insights gained from images, manuals, and videos. The search volume should support autocompletion and suggestions. All raw insight data generated must be stored for later processing. Stock level fields in the product index must update immediately upon changes to prevent purchases of out-of-stock items. Shopping on the go requirements for the chatbot include answering common questions, supporting interaction in all three languages, and replacing the existing FAQ process with a central location for questions and answers. All employees should be able to edit questions and answers, but only senior managers can publish updates. The chatbot must support purchases by providing relevant product information, and product displays must include images and warnings for low or out-of-stock levels. A sample product JSON is provided, showing fields for SKU, name (in English, Spanish, Portuguese), stock level, description, image URI, and alt text (in English, Spanish, Portuguese). The alt text is initially created in English. The first question (41) asks how to complete blank one for an adaptive card in the chatbot, specifically for displaying the product name. The correct approach is to dynamically display the name based on the user's chosen language, using `name.[language]` to retrieve the appropriate language version (e.g., `name.en`, `name.es`, `name.pt`). The second question asks how to complete blank two, which involves displaying a text block with an "attention" color. This is intended for displaying stock levels when they are not "okay" (i.e., low or out of stock). The correct condition for the `when` keyword is `stockLevel != 'OK'`. The third question asks how to complete blank three for the alt text of an image. Similar to the product name, the alt text needs to be dynamically displayed in the user's preferred language. The correct syntax is `image.allText.[language]`, allowing for dynamic retrieval of the alt text in English, Spanish, or Portuguese.