AI PM Technical Fluency Course

A 25-session hands-on course to learn AI/LLM development, using the Anthropic Console for interactive experiments.

Last updated: February 2026 | Main model: Claude Opus 4.5

How It Works

Each session is ~1 hour of hands-on learning. Claude Code serves as your instructor — guiding you through experiments, explaining concepts, and tracking progress.

Running example: You'll build an AI feature for YOUR product throughout the course, applying each concept to something real.

Quick Start

cd path/to/AIPM
claude

Then say:

• "Let's get started" - Run first-time setup (name, product, etc.)
• "Let's do Session 1" - Start a specific session
• "Continue my course" - Resume where you left off
• "Show my progress" - See completion status

Files

File	Purpose
README.md	Course overview (this file)
LESSONS.md	Detailed plans for all 25 sessions
CLAUDE.md	Instructor instructions
user.json	Your profile and product (created at setup)
progress.json	Your progress data
PROGRESS.md	Your progress with checkmarks
MY_PRODUCT.md	Your running example
workspace/.env	API keys

The 25 Sessions

Phase 1: Foundations (Sessions 1-6)

1. LLM Architecture & API Fundamentals
2. Prompt Engineering Patterns
3. Embeddings & Semantic Similarity
4. Vector Databases Deep Dive
5. Document Processing & Chunking
6. Building Your First RAG Pipeline

Phase 2: Agentic AI (Sessions 7-13)

7. Function Calling & Tool Use
8. ReAct Agents & Extended Thinking
9. Memory Systems for Stateful Agents
10. Advanced RAG Patterns
11. Corrective & Autonomous RAG
12. When RAG Fails — Agentic Search Alternatives
13. Voice & Multimodal Agents

Phase 3: Multi-Agent Systems (Sessions 14-19)

14. Multi-Agent Architecture Patterns
15. CrewAI for Role-Based Teams
16. LangGraph for Complex Workflows
17. OpenAI Agents SDK Deep Dive
18. Google ADK for Enterprise Agents
19. MCP & Standardized Tool Calling

Phase 4: Production (Sessions 20-25)

20. Evaluation Frameworks That Matter
21. Safety, Guardrails & Responsible AI
22. RLHF & Alignment Deep Dive
23. System Design for AI Products
24. Fine-tuning vs RAG vs Agentic Search Decision Framework
25. Capstone Integration

Prerequisites

• Access to console.anthropic.com
• An Anthropic API key (add to workspace/.env)