The Modern AI Stack: Introduction
- What This Course Is and What It Isnt
- How We Got Here From Rules to Transformers
- The Modern AI Stack Models Harnesses Agents and Tools
Inside the Model
- What an LLM Actually Is Parameters Weights and Training vs Inference
- LLM vs SLM Open vs Closed Weights
- Tokens The Currency of AI
- Context Windows What Fits What Falls Out and Why It Matters
- Temperature and Sampling Why the Same Prompt Gives Different Answers
- Hallucinations Why Models Make Things Up
Choosing and Running Models
- The Model Marketplace Families Versions and What Your Tools Are Built on
- Token Economics What AI Actually Costs
- The Selection Triangle Performance Cost and Latency
- Hosted APIs vs Running Your Own The Real Tradeoffs
- Demo Running a Local Model
Prompts, Context, and RAG
- Prompt Engineering for Professionals System Prompts Few Shot and Chain of Thought
- From Prompt Engineering to Context Engineering
- RAG Part 1 Embeddings and Vector Search No Math Required
- RAG Part 2 The End to End Pipeline
- RAG vs Fine Tuning vs Long Context Which Lever to Pull When
AI Agents and Agentic AI
- The Agent Loop What Makes an Agent an Agent
- Tool Calling How Models Touch the World
- The Harness The Other Half of the Product
- Subagents Multi Agent Systems and Orchestrators
- Building Subagents Instructions Tools and Delegation
- Human in the Loop Approvals Checkpoints and Guardrails
The Ecosystem: MCP, Frameworks, and Automation
- Why AI Integration Was Broken The NM Problem
- MCP Hosts Clients and Servers
- Demo Connecting an MCP Server
- Orchestration Frameworks LangChain LangGraph and CrewAI
- Skills and Automation Teaching AI Your Workflows
AI Security Essentials
- Prompt Injection The Top Threat to AI Systems
- Overprivileged Agents Governance and Least Privilege
- Identity Runtime Security and Observability