GenAI Beginner Course
Build your first production-ready GenAI application with AgentFlow. This course takes you from "I know what an LLM is" to "I can ship a GenAI app with tools, memory, and evaluation."
What You'll Build
A small engineer-facing assistant that:
- Answers questions using a curated knowledge source
- Uses tools safely (calculator, search, file operations)
- Accepts file or multimodal input
- Returns structured output (JSON)
- Supports thread continuity and memory
- Streams responses to a client
- Includes evaluation and a release checklist
What You'll Learn
| Lesson | Topic | Key Concept |
|---|---|---|
| 1 | Use cases, models, and the LLM app lifecycle | Pick the right use case before building |
| 2 | Prompting, context engineering, and structured outputs | Build reliable outputs with schemas |
| 3 | Tools, files, and MCP basics | Extend the agent with safe tool use |
| 4 | Retrieval, grounding, and citations | Ground answers in real knowledge |
| 5 | State, memory, threads, and streaming | Build conversation-aware applications |
| 6 | Multimodal and client/server integration | Connect to frontends and handle files |
| 7 | Evals, safety, cost, and release | Ship with confidence |
Course Structure
Prerequisites
- Python basics (functions, classes, async/await)
- Comfortable with API request/response formats
- No prior LLM or agent experience needed
Time Commitment
| Component | Time |
|---|---|
| 7 lessons | 30-45 min each |
| Capstone exercise | 1-2 hours |
| Total | ~5-6 hours |
How Each Lesson Is Structured
Every lesson includes:
- Concept — Brief explanation with diagrams
- Example — Complete, runnable AgentFlow code
- Exercise — Try it yourself with guidance
- What you learned — Key takeaways
- Next step — Where to go next
AgentFlow Concepts You'll Master
| Concept | Where It's Used |
|---|---|
| StateGraph | Lesson 1+ |
| Tools and validation | Lesson 3 |
| Structured outputs | Lesson 2, 7 |
| Memory and stores | Lesson 4, 5 |
| Checkpointing | Lesson 5 |
| Streaming | Lesson 5, 6 |
| Client integration | Lesson 6 |
Your Learning Path
Start Here
If you're new to AgentFlow, start with these shared foundations:
- LLM basics for engineers — What LLMs are
- Tokenization and context windows — Why tokens matter
- Prompt patterns cheatsheet — Reliable prompting
Then Continue With Lessons
Start with Lesson 1: Use cases, models, and the LLM app lifecycle
After This Course
After completing this course, you'll be ready for:
- Advanced Course: Agentic product fit and system boundaries
- Production deployment: How-to guides
- Real projects: Build your own GenAI applications
Coming from the Beginner Path?
If you've already completed the Beginner Path, this course goes deeper into the "why" and "when" of GenAI system design. The lessons will feel familiar but with more context.
Ready to start? Begin with Lesson 1: Use cases, models, and the LLM app lifecycle.