AI Bot
Azure AI Services & Cognitive Services Flashcards
Microsoft Azure AI Fundamentals AI-901 Flashcards
Study our Azure AI Services & Cognitive Services flashcards for the Microsoft Azure AI Fundamentals AI-901 exam with 44+ flashcards. View as flashcards, a searchable table, or as a fun matching game.
| Front | Back |
| How can you improve privacy when using Cognitive Services | Use private endpoints customer managed keys and on premise containers where available |
| How do you monitor and evaluate Cognitive Service models in production | Use Azure Monitor Application Insights and model specific metrics |
| How do you secure API keys for Cognitive Services in production | Store in Azure Key Vault and use managed identities |
| How does Azure Bot Service connect to channels | Via channel connectors such as Teams Slack Web Chat and Direct Line |
| How does Form Recognizer differ from generic OCR | It returns structured key value pairs tables and semantic labels tailored to forms |
| Name a common use case for Computer Vision | Image tagging object detection and OCR |
| Name a common use case for Custom Vision | Custom image classification for labels unique to your business |
| Name a common use case for Text Analytics | Sentiment analysis and key phrase extraction |
| What are the four main Cognitive Services categories | back,Vision Speech Language Decision |
| What authentication methods are commonly used with Cognitive Services | API keys and Azure Active Directory |
| What Azure service is best for OCR and image analysis | Computer Vision |
| What Azure service is best for training an image classification model with minimal code | Custom Vision |
| What Azure service provides language understanding and question answering | Azure Language Service |
| What deployment consideration matters for low latency inference | Deploy models in region closest to users or use edge containers |
| What factors affect the decision between using Azure OpenAI and managed Cognitive Services | Need for generative capabilities control cost data privacy and compliance requirements |
| What is a conversational skill in the context of bots | Composable component that adds a domain specific capability like booking or QnA |
| What is a custom model in Azure AI | Model trained on your data for domain specific tasks |
| What is a prebuilt model in Cognitive Services | Model provided by Microsoft ready to use without custom training |
| What is a typical tradeoff with managed services | Less control over model internals but faster deployment and easier maintenance |
| What is Direct Line used for | Programmatic channel to connect custom apps to bots |
| What is the advantage of using the Speech SDK | Low latency real time features and client side audio processing |
| What is the benefit of using multimodal embeddings in Language service | Enable semantic search retrieval and RAG workflows across text and images |
| What is the best choice for a simple FAQ bot | Azure Bot Service with Language service Question Answering |
| What is the difference between Computer Vision and Custom Vision | Computer Vision provides general image APIs Custom Vision allows you to train custom classifiers |
| What is the Document Intelligence name sometimes used for | Advanced Form Recognizer capabilities for documents and receipts |
| What is the role of containers in Azure Cognitive Services | Run services on premise for offline scenarios and data residency |
| What replaced QnA Maker functionality | Language service Question Answering |
| What tool helps design bots with low code | Bot Framework Composer |
| When is it appropriate to fine tune a model | When you need to adapt a large model to domain specific language or tasks |
| When should you build a custom model instead of using a managed service | Build custom model when you need high domain specific accuracy custom features or on premise deployment |
| When should you choose a managed Cognitive Service over a custom model | Choose managed service for faster time to market lower maintenance and built in capabilities |
| Which Cognitive Service supports multimodal features including embeddings and retrieval augmented generation | Azure Language Service |
| Which service converts speech audio to text | Speech to Text |
| Which service converts text to natural sounding speech | Text to Speech |
| Which service detects anomalies in time series data | Anomaly Detector |
| Which service detects faces age emotions and landmarks | Face API |
| Which service extracts entities sentiment and key phrases from text | Text Analytics |
| Which service extracts structured data from forms and invoices | Form Recognizer |
| Which service helps build chatbots and integrates with channels | Azure Bot Service |
| Which service is used for content moderation of images text and videos | Content Moderator |
| Which service manages real time speech translation | Speech Translation |
| Which service provides profanity detection and image moderation | Content Moderator |
| Which service recommends actions personalization and ranking | Personalizer |
| Which service would you use to extract table data from scanned invoices | Form Recognizer Layout and Document extraction |
About the Flashcards
Flashcards for the Microsoft Azure AI Fundamentals exam help students review Azure Cognitive Services and Azure AI concepts, including vision, speech, language, and decision services. The deck focuses on matching common workloads to the right service, such as OCR, image analysis, speech translation, text analytics, chatbots, and form extraction.
Students can use these cards to reinforce key terminology and compare managed services, custom models, prebuilt models, and deployment options. The deck also reviews practical considerations such as authentication, API keys, privacy, monitoring, containers, low-latency inference, and when to choose Azure OpenAI or managed Cognitive Services.
Topics covered in this flashcard deck:
- Cognitive Services categories
- Vision and OCR
- Speech and language
- Bots and QnA
- Forms and documents
- Security and deployment