AI-102 Practice Exam — AI-102: Azure AI Engineer Associate

" target="_blank" rel="nofollow noopener">https://www.examtopics.com/assets/media/exam-media/04271/0000200002.png">

" target="_blank" rel="nofollow noopener">https://img.examtopics.com/ai-102/image199.png">

The scenario outlines three core requirements for the chatbot: support for chit-chat, a knowledge base, and multilingual models; sentiment analysis on user messages; and automatic selection of the best language model. The services in option C directly address every requirement. QnA Maker (now part of Azure AI Language question answering) provides built-in support for custom knowledge bases and preconfigured chit-chat personality datasets, and natively supports multilingual knowledge base deployments. Text Analytics, another component of Azure AI Language, includes sentiment analysis as a core feature, which evaluates user messages to detect positive, negative, or neutral sentiment. Language Understanding (now Conversational Language Understanding, part of Azure AI Language) supports multilingual intent recognition, automatically detects the input language of user messages, and routes queries to the most appropriate trained language model for processing, fulfilling the automatic model selection requirement. All three services integrate seamlessly to meet all specified requirements for the chatbot. Option Analysis: A. This option is incorrect. While QnA Maker and Language Understanding address the knowledge base, chit-chat, and multilingual model requirements, and Dispatch is used to route queries between multiple LUIS apps and QnA Maker instances, this option lacks Text Analytics, which is required to perform sentiment analysis on user messages. No service included here provides sentiment analysis functionality, so it fails to meet the second core requirement. B. This option is incorrect. Translator provides text translation, Speech enables speech-to-text and text-to-speech functionality, and Dispatch routes queries between conversational AI services. None of these services support knowledge base hosting, chit-chat, or sentiment analysis, so this option fails to meet multiple core requirements of the scenario. C. This option is correct. QnA Maker provides the required knowledge base and chit-chat support with native multilingual deployment capabilities. Text Analytics delivers the required sentiment analysis functionality for user messages. Language Understanding supports multilingual intent detection, automatically identifies the input language of user queries, and selects the best trained language model to process the request, fulfilling all three specified requirements for the chatbot. D. This option is incorrect. While Text Analytics provides sentiment analysis, this option lacks QnA Maker for knowledge base and chit-chat support, and lacks Language Understanding for automatic language model selection. Translator only performs text translation and does not handle routing to appropriate language models, so this option fails to meet the first and third core requirements. Key Concepts: 1. Azure AI Language Service Components: The Azure AI Language service is a unified set of natural language processing capabilities that includes question answering (formerly QnA Maker), conversational language understanding (formerly LUIS), and Text Analytics features. These components are designed to integrate seamlessly for conversational AI workloads, and are a core domain of the AI-102 certification. 2. Conversational AI Requirement to Service Mapping: For chatbot implementations, specific Azure AI services map directly to common requirements: question answering for knowledge bases and chit-chat, Text Analytics for sentiment analysis and text insight extraction, and conversational language understanding for intent classification and language-aware routing. 3. Multilingual Chatbot Functionality: Native multilingual support in Azure AI Language services allows chatbots to automatically detect input language and use pre-trained language-specific models without requiring a separate translation layer, reducing implementation complexity and improving response performance. References: Choose the right Azure AI services for conversational AI, Question answering in Azure AI Language, https://learn.microsoft.com/en-us/azure/ai-services/language-service/question-answering/overview

The scenario requires extracting structured top-level fields from English receipts with minimal development effort. Azure Form Recognizer (now branded Azure AI Document Intelligence) provides a prebuilt receipt model that is pre-trained to recognize and extract common receipt fields including vendor name, transaction total, transaction date, line items, and more out of the box. No custom model training or complex text parsing logic is required to use this prebuilt model, directly meeting the minimal development effort requirement for the expense report receipt processing use case. This aligns with AI-102 domain knowledge covering appropriate Azure AI service selection for document processing workloads. Option Analysis: A. Custom Vision is incorrect. Custom Vision is an Azure AI service for building custom image classification and object detection models, designed for use cases like identifying product types in images or detecting defects in manufacturing parts. It does not support structured data extraction from receipts or other document types, so it cannot address the scenario requirements. B. Personalizer is incorrect. Personalizer is a reinforcement learning service used to deliver personalized content recommendations and user experiences, such as suggesting products to customers or optimizing UI element placement. It has no functionality related to document processing or text/data extraction from receipts, making it irrelevant to the scenario. C. Form Recognizer is correct. Form Recognizer's prebuilt receipt model eliminates the need for custom development or model training to extract required fields (vendor, transaction total) from English receipts, directly satisfying both the functional and development effort requirements of the scenario. D. Computer Vision is incorrect. The general Azure Computer Vision service includes optical character recognition (OCR) capabilities to extract raw unstructured text from images of receipts, but it does not map that raw text to structured fields like vendor name or transaction total. Implementing the required extraction with Computer Vision would require building custom text parsing and field mapping logic, which adds significant development effort, violating the scenario's minimal development requirement. Key Concepts: 1. Form Recognizer Prebuilt Models: Pre-trained, ready-to-use models for common document types (receipts, invoices, ID documents) that extract structured fields without custom training, reducing development overhead for common document processing use cases. 2. OCR vs Structured Document Extraction: General OCR services extract raw unstructured text from images, while dedicated document intelligence services like Form Recognizer add context-aware parsing to map text to predefined structured key-value pairs for direct business use. 3. Azure AI Service Use Case Alignment: A core AI-102 competency requiring matching workload requirements to the most appropriate Azure AI service, prioritizing out-of-the-box capabilities to minimize development effort where possible. References: What is Azure AI Document Intelligence (formerly Form Recognizer)?, https://learn.microsoft.com/en-us/azure/ai-services/document-intelligence/overview?view=doc-intel-4.0.0 Prebuilt receipt model for Azure AI Document Intelligence

" target="_blank" rel="nofollow noopener">https://www.examtopics.com/assets/media/exam-media/04271/0001000001.png">