Azure AI Fundamentals | Yael Bernstein

Key Takeaway:

A solid understanding of what AI means in Azure’s context, and how different workloads (vision, speech, text, generative AI) are supported.

Ability to distinguish between ML techniques (supervised, unsupervised, deep learning) and understand how to evaluate them.

Awareness of responsible AI practices and ethical considerations that must be part of AI solution design.

AI-900: Azure AI Fundamentals — Day 1 & Day 2 (April) Summary

📆 Day 1 – Core Concepts & Machine Learning Fundamentals

Introduction to AI & Azure AI Fundamentals
What artificial intelligence is; what kinds of workloads AI enables (prediction, classification, recommendation, anomaly detection). Responsible AI principles (fairness, reliability, privacy, transparency)
Azure AI Services Overview
Basic overview of Azure services relevant to AI: Azure Machine Learning, Cognitive Services, Azure OpenAI Service, etc.
Machine Learning Fundamentals
Concepts such as supervised vs. unsupervised learning, training / validation / testing datasets, model evaluation, features and labels, overfitting/underfitting. Possibly some introduction to deep learning and transformer architecture at a high level.
Use-Cases & Scenarios
Real world examples of how ML is applied (e.g. for forecasting, classification, clustering). Understanding when to use ML, what data is needed, what performance metrics matter.

📆 Day 2 – Vision, NLP, Generative AI & Tools

Computer Vision Workloads: Image classification, object detection, OCR (Optical Character Recognition), image analysis, possibly face detection/analysis. Azure services like Azure Vision, or Computer Vision API.
Natural Language Processing (NLP): Key phrase extraction, entity recognition, sentiment analysis, speech recognition and synthesis, text translation. Tools like Azure Language Services, Speech Services.
Generative AI: What generative AI is (text generation, prompts, LLMs), use cases. Azure services for gen AI (OpenAI, etc.). Also examining how to use generative models responsibly.
Responsible AI and Ethical Considerations: More depth on guiding principles: fairness, safety, security, inclusivity, accountability. How Azure implements tools or services / resources to help (policies, oversight, monitoring).
Hands-On / Tools & Labs: Using Microsoft Learn labs or sandboxes; exploring Azure Machine Learning Studio; experimenting with cognitive services; possibly doing small tasks in vision/NLP/generative AI contexts.

👉 Think of it this way:
Azure AI = 3 layers