SESSION 36: ENTERPRISE AI & LLM FUNDAMENTALS
Topics:
- Generative AI & LLMs (GPT-4o, Llama 3) mechanics
- Tokens, Context Windows, Temperature, and Hallucination controls
- Prompt Engineering for admins (System Prompts, Few-shot)
- Microsoft Copilot ecosystem across Windows, M365, & Azure
PRACTICAL LAB:
- Lab 1: Craft prompts to analyze system errors and draft RCA reports
SESSION 37: SCRIPT GENERATION WITH COPILOT
Topics:
- Generating PowerShell, Python, and Bicep scripts with AI
- Debugging syntax/logic errors and code optimization
- Auditing AI-generated code for security flaws
PRACTICAL LAB:
- Lab 1: Use Copilot to write, debug, and optimize complex PowerShell scripts
SESSION 38: AI-POWERED DIAGNOSTIC TRIAGE
Topics:
- Parsing Event Viewer XML & server logs using LLMs
- AI-assisted outage triage (High CPU, crashes, auth failures)
PRACTICAL LAB:
- Lab 1: Feed unformatted error logs to Copilot for root cause analysis
SESSION 39: AZURE AI FOUNDRY & OPENAI
Topics:
- Deploying Azure OpenAI instances and models (GPT-4o)
- API Key management, endpoints, and Content Safety filters
- Token quotas, rate limits, and cost management
PRACTICAL LAB:
- Lab 1: Deploy a GPT-4o model in Azure AI Foundry & configure safety filters
SESSION 40: LOCAL LLM DEPLOYMENT & ADMINISTRATION
Topics:
- Running local open-source LLMs (Llama 3, Mistral) via Ollama
- Hardware requirements (GPU VRAM, CPU, Quantization)
- Deploying local Open WebUI interfaces for offline privacy
PRACTICAL LAB:
- Lab 1: Install Ollama and run a private LLM with Open WebUI interface
SESSION 41: ENTERPRISE AI AGENTS & RAG
Topics:
- Retrieval-Augmented Generation (RAG) concepts
- Vector Databases (Azure AI Search) overview
- Building internal IT Helpdesk Knowledge Bots
PRACTICAL LAB:
- Lab 1: Build a custom RAG Chatbot using company IT documentation
SESSION 42: AI SECURITY & GOVERNANCE
Topics:
- Prompt Injection attacks and prevention
- Data Loss Prevention (DLP) for AI using Microsoft Purview
- Responsible AI enterprise policies
PRACTICAL LAB:
- Lab 1: Configure Purview DLP rules to prevent sensitive data leaks into AI tools
🏆 MODULE PROJECT: DEPLOY AN AI-POWERED IT HELP-DESK & AUTOMATION PIPELINE