Designing and Implementing AI Solutions with Azure and C#

Define artificial intelligence and explore its everyday use in home, phone, car, ChatGPT, and beyond, including visual perception, text analytics, NLP, speech, and Azure AI services for IoT anomaly detection.

Define data science, machine learning, and AI, highlighting data processing, statistics, and predictive models. See real-world examples from flight delays to healthcare to illustrate how these pillars power AI solutions.

Explore Azure machine learning capabilities, including automated ML, designer, data and compute for training and deploying models as web services; leverage OpenAI, prompts, and cognitive search to build intelligent apps.

Provision Azure AI services and use language detection via REST and SDK in C# or Python. Configure endpoint and key to access text analytics.

Implement multi-layer security for Azure AI services using authentication, key management, and network controls; rotate production and development Cognitive Services keys and protect access with Azure Key Vault.

Configure Azure Key Vault to store secrets and grant access to a service principal, then update your C# or Python app to fetch the secret with Azure identity.

Monitor cost, utilization, and potential issues in Azure AI services. Use the Azure Pricing calculator, cost analytics, and alerts to track cognitive services usage with matrix charts and activity logs.

Explore Azure AI containers to run cognitive services on premises with Docker. Leverage hybrid data control, scalable deployment, and low latency across Azure Container Instance, Kubernetes, and edge.

Explore Azure AI Vision's image analysis features, including description, captions, tags, and object detection. Also detect people, remove backgrounds, or use OCR and facial recognition.

Devise an image analysis workflow in C# using the Azure Vision SDK, authenticating with endpoint and key, and retrieving captions, dense captions, tags, and object detection results with bounding boxes.

Practice object detection and locating objects in images by drawing bounding boxes and saving annotated results, then detect people with confidence scores and perform background removal and foreground matting.

Explore image classification with Azure Custom Vision, train a model to label images (apple, banana, orange), and publish predictions using separate training and prediction resources for object detection.

Classify fruit images by building a custom image classification model with Azure Vision, training on labeled fruit images, storing assets in a blob-enabled storage account, and evaluating with test images.

Explore object detection in computer vision, training models with Azure Custom Vision to predict object classes and localize them with bounding boxes, using training and prediction resources.

Build an Azure Custom Vision object detector that identifies apples, bananas, and oranges, creating training and prediction resources, tagging images, and training the model with C#.

Test a trained fruit detector model with a sample image, publish the model to the custom vision service, and run the C# app to produce a labeled output JPEG.

Azure AI vision services detect and analyze faces, return bounding boxes, IDs, head pose, and facial landmarks, and cover limited access policies and responsible AI ethics for facial recognition.

Explore how Azure Video Indexer extracts OCR text, speech transcription, topics, sentiment, and scene segmentation, plus custom insights via the Video Indexer API.

Learn to use Azure language to detect language, extract key phrases, analyze sentiment, and link entities with Wikipedia, via REST or SDK.

Clone the lab repository, configure Python text analytics, and detect language, sentiment, key phrases, and linked entities from hotel reviews using the Azure Language Text SDK.

Explore creating a knowledge-base driven question answering service in Azure language, migrate from QnA Maker, publish a Rest endpoint, and support multi-turn conversations with follow-up prompts.

Create a knowledge base for a question answering bot using Azure language services, custom question answering, and imports from Microsoft Learn FAQ and chit chat; train, test, and deploy.

Deploy a knowledge base with a REST endpoint and prediction URL, then implement a C# client using Azure language services to submit questions and return answers with confidence.