Azure Al Vision
Tools for processing, analyzing, and interpreting visual content (images & video)
Facial Detection and Analysis
can be used to detect and verify faces
Optical Character Recognition (OCR)
extracts text from images
Azure Al Video Indexer
provides video analysis including speech-to-text transcription and tone assessment
Spatial Analysis
detects presence and movement of people in real-time video
Azure Al Document Intelligence
Services that allow us to understand, analyze, and extract meaningful data from documents.
Documents include: PDFs, Images, emails, scanned texts, etc.
-More than just OCR
-Financial services & Legal models
-Invoice Extraction
- Bank Statement
- Credit card & Check reading
-Personal Identification
-Tabular data extraction
-Keyword extraction
Azure Al Search
Enables Al systems to quickly & efficiently search for relevant data.
-Often used for RAG systems.
-Typically employes “Embeddings” to convert data to something AI inherently understands.
-Full text & vector search.
Azure Al Language
Enables Natural Language processing, Speech to Text, and Text to Speech capabilities.
Speech Recognition:
The Al “Listens”
Audio to Text
Speech Synthesis:
The Al “Talks”
Azure OpenAl Service
Azure OpenAl Service allows you to consume OpenAI LLM models from your Azure account.
APIs for popular foundation models, such as:
GPT-5 Series models
GPT-4 Series
GPT-oss-120b,20b
Sora 2
Consuming Azure Al Services
Azure provides three ways to consume their Al services: • Web GUIs (such as Microsoft Foundry) • SDKs • REST API endpoints. We’ll mostly use SDKs and REST API endpoints in this class to integrate these services into code.
Best for Visual Configuration and Testing
Use the graphical interface for a point-and-click experience to provision services, test models in playgrounds, and monitor performance without writing code.
The Preferred Choice for Streamlined Development
Leverage language-specific libraries (Python, NET, Java, Node.js) that abstract complex HTTP requests into simple, predefined classes and methods.
Maximum Flexibility and Language Independence
Use standard HTTP methods (GET, POST) and JSON responses to integrate Al into any application or script that can make web requests.