Home Courses Azure AI
Data Science & AI ⚡ Microsoft Azure AI · OpenAI · Semantic Kernel ✔ 3 Levels Available

Microsoft Azure AI
Full Programme

An enterprise-grade, 48-hour hands-on curriculum engineered for AI engineers and machine learning practitioners. Master Azure AI Foundry, Azure OpenAI Service (GPT-4o), Document Intelligence & Cognitive APIs, Azure Machine Learning Workspaces & Pipelines, Azure AI Search with Semantic Ranker, Phi-3 Small Language Model adaptation, Azure AI Agent Service, AutoGen, Semantic Kernel, and Prompt Flow LLMOps aligned with Microsoft Certified: Azure AI Engineer Associate (AI-102) and Azure Data Scientist (DP-100) standards.

🎯
Format
Theory + Lab/Assignments (48 Hours)
🏆
Certificate
Arich Certified Azure AI Engineer
Enrol Now →
Course Code
AZURE-AI
Click a Level to View its Syllabus
L1
AZ-AI-FND · Associate Level
6 Modules · OpenAI, AI Search & Prompt Flow
›
L2
AZ-AI-PRO · Professional Level
4 Modules · Semantic Kernel & AutoGen
›
L3
AZ-AI-EXP · Enterprise AI Architect
4 Modules · Fine-Tuning & Zero-Trust AI
›
📄 Azure AI — Associate Level (AI-102 Aligned — 48 Hours)
L1
AZURE-AI-FND
Azure AI Foundry, OpenAI, Cognitive Services, Azure ML & AI Search RAG
01
Azure AI Foundry & OpenAI Essentials
▼
1.1 Cloud AI & GenAI Concepts: Overview of Generative AI on Azure, Azure AI Studio & AI Foundry architecture
Azure OpenAI Service: Provisioning, Model Deployments (GPT-4o, GPT-4o-mini, DALL-E 3, Whisper)
Prompt Engineering on Azure: System Messages, Few-Shot Prompting, Temperature & Top-P parameter tuning
Azure AI Content Safety: Content Filtering, Blocklists, Jailbreak Detection & Protected Material Filters
Hands-on Lab: Deploy GPT-4o on Azure OpenAI Service, author system instructions, and configure custom Content Safety thresholds
Duration: Theory: 4.0 Hours | Lab/Assignments: 4.0 Hours | Total: 8.0 Hours
02
Specialized Azure AI & Cognitive Services
▼
2.1 Specialized Azure AI Services Architecture: Multi-Service Cognitive Account setup and authentication
Azure AI Vision: Image Analysis 4.0, Optical Character Recognition (OCR), Face Analysis & Spatial Analysis
Azure AI Language: Named Entity Recognition (NER), Sentiment Analysis, Conversational Language Understanding (CLU), Text Summarization
Azure AI Speech: Real-Time Speech-to-Text, Neural Text-to-Speech Custom Voices & Speech Translation
Azure Document Intelligence (formerly Form Recognizer): Prebuilt Receipt/Invoice Models, Layout Analysis & Custom Extraction Models
Hands-on Lab: Build an Automated Financial Invoice Processing Pipeline using Azure Document Intelligence, Language Sentiment & Azure Functions
Duration: Theory: 4.0 Hours | Lab/Assignments: 4.0 Hours | Total: 8.0 Hours
03
Azure Machine Learning (AML) Workspace & Data Pipelines
▼
3.1 Classical ML & Managed Data Pipelines: Azure Machine Learning Workspaces, Compute Instances & Compute Clusters
Data Assets, Managed Datastores (Azure Blob, ADLS Gen2) & Azure ML Feature Store
Automated ML (AutoML): Regression, Classification & Time-Series Forecasting with Explainable AI
Azure ML Pipelines: Component-based ML workflow orchestration, Data Prep, Model Training & Hyperparameter Tuning (Sweep Jobs)
Hands-on Lab: Train, Evaluate, and Register a Custom Machine Learning Model using Azure AutoML and ML Pipelines with Real-Time Endpoint Deployment
Duration: Theory: 4.0 Hours | Lab/Assignments: 4.0 Hours | Total: 8.0 Hours
04
Azure AI Search & Enterprise RAG
▼
4.1 Knowledge Retrieval & Vector Search Architecture: Azure AI Search indexing pipeline (Indexers, Skillsets, Crack Document, Vectorization)
Vector Indexing: HNSW (Hierarchical Navigable Small World) algorithms and Cosine/DotProduct metric configuration
Hybrid Search: Combining BM25 keyword matching with dense text-embedding-3 vectors
Semantic Ranker: Deep learning re-ranking models for high-precision relevant answer generation
Integrated Vectorization with Azure OpenAI text-embedding models and automated chunking
Hands-on Lab: Construct Enterprise Retrieval-Augmented Generation (RAG) Solution grounded on Azure AI Search with Semantic Ranker & GPT-4o
Duration: Theory: 4.0 Hours | Lab/Assignments: 4.0 Hours | Total: 8.0 Hours
05
Fine-Tuning & Model Catalog Customization
▼
5.1 SLM & LLM Adaptation: Azure AI Studio Model Catalog (Phi-3 Mini/Medium, Llama 3.1, Mistral, Command R+)
Supervised Fine-Tuning (SFT) and Parameter-Efficient Fine-Tuning (PEFT/LoRA) on Small Language Models (SLMs)
Model Optimization & Acceleration with ONNX Runtime & Olive Tooling for Edge/Cloud Inference
Azure ML Hyperparameter Tuning & Automated LLM Evaluation Benchmarking
Hands-on Lab: Fine-Tune Microsoft Phi-3 SLM on Custom Domain Dataset in Azure AI Studio and Deploy with ONNX Runtime Acceleration
Duration: Theory: 4.0 Hours | Lab/Assignments: 4.0 Hours | Total: 8.0 Hours
06
Agentic Workflows, Azure AI Agent Service & Prompt Flow MLOps
▼
6.1 Agentic Workflows & Multi-Agent Systems: Azure AI Agent Service, AutoGen, Semantic Kernel Framework, Bing Search Tooling, Code Interpreter
7.1 Production Governance & Evaluation: Prompt Flow development, DAG flows, Custom Evaluation Metrics (Groundedness, Relevance, Fluency)
Responsible AI Dashboard: Fairness, Error Analysis & Interpretability (Shapley Explanations)
Production MLOps: Azure DevOps & GitHub Actions CI/CD for AI Models, Automated Evaluation Gates, Telemetry & Cost Management
Capstone Project: End-to-End Enterprise Multi-Agent Copilot Platform with Semantic Kernel, Azure AI Search RAG, Prompt Flow & CI/CD Evaluation
Duration: Theory: 4.0 Hours | Lab/Assignments: 4.0 Hours | Total: 8.0 Hours
📄 Azure AI — Professional Level (AZ-AI-PRO — Multi-Agent & MLOps)
L2
AZ-AI-PRO
Semantic Kernel, Multi-Agent AutoGen, Advanced AI Search & Prompt Flow CI/CD
01
Enterprise Semantic Kernel Framework & Native Plugins
▼
Semantic Kernel Core Architecture: Kernel initialization, Native Plugins, Semantic Functions & Memory Connectors
Planner Engines: Stepwise Planner, Handlebars Planner & Function Calling Orchestration
Context Variable Passing, Kernel Filters, Pre/Post-Invocation Hooks & Streaming Responses
Hands-on Lab: Build an Enterprise Data Copilot using Semantic Kernel with Custom SQL & REST API Native Plugins
02
Multi-Agent Swarms with Microsoft AutoGen on Azure
▼
AutoGen Multi-Agent Architecture: ConversableAgent, UserProxyAgent & GroupChatManager
Hierarchical Agent Collaboration: Manager, Planner, Coder, Reviewer & Critic Agents
Sandboxed Code Execution using Azure Container Instances (ACI) & Docker
Hands-on Lab: Construct an Autonomous Multi-Agent Software Development Swarm on Azure with AutoGen & Git PR Automation
03
Advanced Vector Search, Chunking & Semantic Ranker Tuning
▼
Hierarchical & Semantic Document Chunking: Overlap windows, Markdown/PDF structure preservation
Multi-Vector Search, Cross-Lingual Vector Search & Multi-Modal Embedding integration
Azure AI Search Index Optimization: Sharding, Replica configuration, Filtering with OData expressions
Hands-on Lab: Deploy Ultra-High-Scale Distributed Azure AI Search Index with Custom Integrated Vectorizer & Semantic Ranker
04
Automated Prompt Flow CI/CD & LLMOps Telemetry
▼
Prompt Flow DAG Orchestration: Standard Flows, Chat Flows & Evaluation Flows
Automated Benchmark Testing with Azure AI Evaluation SDK (Groundedness, Relevance, Coherence, Similarity)
Azure DevOps & GitHub Actions CI/CD Pipeline integration with Automated Quality Deployment Gates
Hands-on Lab: Production Prompt Flow Deployment Pipeline with Automated Regression Evaluation & Application Insights Tracing
📄 Azure AI — Enterprise Architect Level (AZ-AI-EXP — Fine-Tuning & Zero-Trust AI)
L3
AZ-AI-EXP
Enterprise Model Adaptation, Zero-Trust AI Security, Graph-RAG & FinOps
01
High-Performance Model Adaptation (Phi-3 / Llama 3.1 Fine-Tuning)
▼
Supervised Fine-Tuning (SFT) & Direct Preference Optimization (DPO) on Azure Machine Learning Compute Clusters
QLoRA, DeepSpeed & FlashAttention Optimization for Multi-GPU Distributed Training on Azure NC/NDv4 VMs
Model Distillation & Compression: Converting Heavy LLMs into Fast Edge-Ready Phi-3 variants
Hands-on Lab: Enterprise Domain Fine-Tuning & DPO Alignment on Azure ML with Distributed Multi-GPU Clusters
02
Zero-Trust AI Security, Private Endpoints & Content Safety
▼
Zero-Trust AI Architecture: Azure Virtual Network Isolation, Private Endpoints for Azure OpenAI & AI Search
Customer-Managed Encryption Keys (Azure Key Vault CMEK) for Embeddings, Storage & Search Indexes
Azure AI Content Safety: Prompt Shield against Direct & Indirect Prompt Injections, PII Redaction
Hands-on Lab: Build a Zero-Trust GenAI Security Perimeter with Private Link, Key Vault CMEK & Automated Prompt Shield Filters
03
Cognitive Graph-RAG on Azure & Agentic Memory
▼
Graph-RAG Architecture: Combining Knowledge Graphs (Azure Cosmos DB Gremlin / Neo4j) with Azure AI Search
Multi-Hop Reasoning over Complex Enterprise Documents & Entity Relationships
Agentic Long-Term Memory Architectures using Azure Cosmos DB & Azure Cache for Redis
Hands-on Lab: Construct an Enterprise Graph-RAG Intelligence Engine on Azure with Knowledge Graphs & GPT-4o
04
Cloud AI FinOps, Observability & Enterprise Capstone
▼
Azure OpenAI Provisioned Throughput Units (PTU) vs Pay-As-You-Go Capacity Planning
Prompt Caching, Token Optimization & Azure Cost Management AI Allocation Tags
Full-Stack Observability: OpenTelemetry, Application Insights, GenAI Latency SLAs & Token Tracking
Capstone Project: Production Deployment of an Enterprise Generative AI & Multi-Agent Platform on Azure with Full MLOps Telemetry