L16. Azure AI and Machine Learning Services Overview
Video generating
Check back soon for the video lesson on Azure AI and Machine Learning Services Overview
Azure AI services enable developers to add intelligence to applications without deep ML expertise. The AZ-900 exam tests Azure AI Services, Azure Machine Learning, and the responsible AI principles.
Azure AI Services (Cognitive Services)
Azure AI Services (formerly Azure Cognitive Services) are pre-built AI APIs that add vision, speech, language, and decision capabilities to applications without building models from scratch. Key service categories: Vision:
- Computer Vision: analyze images and video, extract text (OCR)
- Face API: detect and recognize faces
- Custom Vision: train a custom image classifier
- Speech to Text: transcribe audio to text
- Text to Speech: synthesize natural-sounding audio from text
- Speech Translation: real-time speech translation
- Language Understanding (LUIS): extract intents and entities from natural language
- Azure OpenAI Service: access GPT-4 and other OpenAI models through Azure
- Translator: translate text between 100+ languages
- Anomaly Detector: identify unusual patterns in time-series data
- Content Moderator: detect offensive content in text, images, video
Azure Machine Learning
Azure Machine Learning is a full platform for building, training, deploying, and monitoring custom machine learning models. Key components:
- Azure ML Studio: web-based IDE for ML workflows
- Automated ML: automatically selects the best algorithm and hyperparameters
- Designer: drag-and-drop model building (no code)
- MLOps: version control, CI/CD pipelines for ML models
Responsible AI
Microsoft publishes six principles for responsible AI that the AZ-900 exam tests:
| Principle | Description |
|---|---|
| Fairness | AI should treat all people fairly |
| Reliability and Safety | AI should perform reliably and safely |
| Privacy and Security | AI should be secure and respect privacy |
| Inclusiveness | AI should empower everyone |
| Transparency | AI should be understandable |
| Accountability | People should be accountable for AI systems |
Azure Bot Service
Azure Bot Service enables building, deploying, and scaling intelligent bots for websites, apps, Teams, and other channels. Exam tip: Azure AI Services = pre-built AI APIs requiring no ML expertise. Azure Machine Learning = custom model building platform. Azure OpenAI = enterprise access to GPT-4 and DALL-E through Azure.
- ✓Azure AI Services provide pre-built AI APIs (vision, speech, language, decision) without needing ML expertise
- ✓Azure Machine Learning is the full platform for building and training custom ML models
- ✓Azure OpenAI Service provides enterprise access to GPT-4 and DALL-E through the Azure platform
- ✓The six Responsible AI principles: Fairness, Reliability, Privacy, Inclusiveness, Transparency, Accountability
- ✓Automated ML automatically selects the best algorithm for your dataset without manual tuning
1. A developer wants to add the ability to detect objects in images to their application without training a machine learning model. Which Azure service is most appropriate?
2. Which Microsoft Responsible AI principle states that AI systems should be understandable and that users should be able to understand how decisions are made?
Recommended: Pluralsight
This free course covers the theory. Pluralsight adds guided video paths, hands-on Azure labs, and timed practice exams to help you pass AZ-900 with confidence.