Azure AI and Machine Learning Services
Key Takeaways
- Azure AI Services (formerly Cognitive Services) provides pre-built AI models for vision, speech, language, and decision-making without requiring ML expertise.
- Azure Machine Learning is a platform for building, training, and deploying custom machine learning models.
- Azure OpenAI Service provides access to GPT-4, DALL-E, and other OpenAI models within Azure's security and compliance framework.
- Azure Bot Service enables building conversational AI bots that interact with users across channels (Teams, web, Slack, etc.).
- For the AZ-900, know WHAT each AI service does and WHEN to use it — not how to build models.
Azure AI and Machine Learning Services
Quick Answer: Azure AI Services = pre-built AI (vision, speech, language). Azure Machine Learning = custom ML model platform. Azure OpenAI = GPT-4 and DALL-E in Azure. Azure Bot Service = conversational AI bots.
Azure AI Services (Formerly Cognitive Services)
Azure AI Services provides pre-built AI models that developers can integrate into applications without needing machine learning expertise:
Vision
| Service | What It Does |
|---|---|
| Computer Vision | Extract text (OCR), identify objects, describe images, detect faces |
| Custom Vision | Build custom image classification and object detection models |
| Face API | Detect and recognize faces, identify emotions, verify identity |
| Form Recognizer | Extract text, key-value pairs, and tables from documents |
Speech
| Service | What It Does |
|---|---|
| Speech-to-Text | Convert spoken audio into text |
| Text-to-Speech | Convert text into natural-sounding spoken audio |
| Speech Translation | Real-time speech-to-speech translation |
| Speaker Recognition | Identify and verify speakers by their voice |
Language
| Service | What It Does |
|---|---|
| Language Understanding (LUIS) | Extract intent and entities from natural language text |
| Text Analytics | Sentiment analysis, key phrase extraction, language detection |
| Translator | Real-time text translation in 100+ languages |
| Question Answering | Build FAQ bots from existing documentation |
Decision
| Service | What It Does |
|---|---|
| Anomaly Detector | Identify anomalies in time-series data |
| Content Moderator | Detect potentially offensive or unwanted content |
| Personalizer | Deliver personalized content and experiences |
Azure Machine Learning
Azure Machine Learning is a cloud-based platform for building, training, and deploying custom machine learning models.
Key capabilities:
- Automated ML (AutoML) — Automatically selects algorithms and tunes hyperparameters
- Designer — Visual drag-and-drop interface for building ML pipelines
- Notebooks — Jupyter notebooks for code-first data science
- MLOps — CI/CD for machine learning models
- Model registry — Version and manage trained models
Azure OpenAI Service
Azure OpenAI Service provides access to OpenAI's powerful models (GPT-4, GPT-4o, DALL-E, Whisper) within Azure's enterprise-grade security, compliance, and regional availability.
Use cases:
- Content generation (text, summaries, translations)
- Code generation and assistance
- Conversational AI and chatbots
- Image generation (DALL-E)
- Audio transcription (Whisper)
Azure Bot Service
Azure Bot Service provides an integrated environment for building conversational AI bots:
- Multi-channel — Deploy to Teams, Slack, web, Facebook, Twilio, etc.
- Integration — Connects with Azure AI Services for NLP capabilities
- Bot Framework SDK — Build bots with C# or JavaScript
- Power Virtual Agents — Build bots with a no-code visual designer
On the Exam: Know the purpose of each AI service — you will not be tested on how to build models. If a question mentions "pre-built AI," the answer is Azure AI Services. If it mentions "custom models," the answer is Azure Machine Learning.
Which Azure service provides pre-built AI models for vision, speech, and language without requiring machine learning expertise?
Which Azure service would you use to build and train CUSTOM machine learning models?
Which service provides access to GPT-4 and DALL-E within Azure's security framework?