I wanted to build the kind of open, structured roadmap I wish I had when I started learning AI development.
Hi everyone 👋
I’m Jaime Lucena, a Generative AI Engineer who’s been building a complete, open-source learning path for anyone who wants to understand — and build — modern AI systems.
Over the past months, I’ve been publishing a series of GitHub repositories that guide you from the absolute fundamentals of Python to production-ready AI agents with LangChain, LangGraph, CrewAI, FastAPI, and Streamlit.
This post walks you through that journey — what each repository teaches, how they connect, and why I believe this structured approach can help you master the real skills behind modern AI engineering.
⸻
🚀 The Goal
There’s an ocean of tutorials out there — but most are fragmented, outdated, or skip the “why” behind the code.
So I decided to create something more practical, modern, and progressive.
Each repo in this series builds directly on top of the previous one.
By the end, you’re not just calling APIs — you’re orchestrating agents, managing memory, and deploying full AI apps.
⸻
🧩 The Learning Path
01 — Python Fundamentals for LangChain
The Python you actually need for building AI applications.
Instead of covering everything Python can do, this repo focuses on the 20% of the language that gives you 80% of the power — the parts directly relevant to working with LLMs and frameworks like LangChain.
You’ll find 9 concise Jupyter notebooks covering:
• Syntax, data types, and functions
• Decorators, OOP, and debugging
• Context managers and environment variables
• Type hints and Pydantic models
Each notebook builds on the previous one, creating a foundation that’s light, modern, and directly applicable to AI work.
🧠 Goal: Learn just enough Python to build, not to memorize.
⸻
02 — LangChain Beginners
A complete, hands-on guide to mastering LangChain fundamentals.
Once you’re confident with Python, this second module introduces the core architecture of LangChain — using its modern LCEL syntax.
It contains 15 structured notebooks, covering everything from prompt templates to RAG and vector databases:
• Build reusable PromptTemplate and ChatPromptTemplate
• Understand and compose Runnable chains
• Learn retrieval-augmented generation (RAG)
• Use vector stores (Chroma / FAISS) for semantic search
• Implement retrievers and similarity search
Each concept is demonstrated through working examples — so instead of theory, you’re constantly seeing how LLM logic connects in code.
🧩 Goal: Move from using LLMs to building your own modular pipelines.
⸻
03 — Agents & Apps Foundations
Learn to build and orchestrate AI agents, manage memory, and deploy them as real applications.
This is where everything comes together — LangGraph, CrewAI, FastAPI, and Streamlit.
Across 7 notebooks, you’ll go from understanding agent theory to deploying a full AI-powered app:
• What AI agents really are (and what they aren’t)
• How LangGraph builds agent workflows with memory
• How CrewAI coordinates multiple agents (e.g. Researcher + Writer)
• Clean architecture patterns for real-world AI apps
• Building and serving APIs with FastAPI
• Creating interactive UIs with Streamlit
By the end, you’ll have a fully functional pipeline — an AI backend with memory and logic, connected to a live frontend.
🧠 Goal: Bridge the gap between “AI experiments” and deployable AI systems.
🧩 How Everything Connects
What makes this series special is not just the code — it’s the structure.
Each repository connects naturally to the next one.
Instead of jumping between random tutorials, you move step by step through the actual stages of becoming a Generative AI Engineer.
| Phase | Repository | Focus |
|---|---|---|
| 🐍 1 | 01-python-fundamentals | The Python you really need for AI work — concise, modern, and hands-on |
| 🔗 2 | 02-langchain-beginners | Building structured, modular LangChain apps with LCEL and RAG |
| 🤖 3 | 03-agents-and-apps-foundations | Orchestrating AI agents and deploying full applications |
By following this progression, you’ll build the full mental model of modern AI systems:
From syntax → to chains → to agent orchestration
From local notebooks → to backend APIs → to real UIs
From theory → to production-ready implementation
This isn’t just about learning a library — it’s about learning how to think in systems.
🧱 Designed for Builders, Not Viewers
Most AI tutorials today teach you how to use tools.
But this series focuses on teaching you how to build with them.
That’s why every concept is:
Tied to code you can run
Explained in context, not isolation
Modular, so you can expand and reuse it later
You’re not memorizing — you’re building reusable mental models and project templates that mirror how real AI engineers work.
🌍 The Vision
My goal with this learning path is simple:
to make the Generative AI engineering discipline more accessible, structured, and practical.
When I started learning about LangChain, RAG, and AI agents, I realized how fragmented the learning experience was.
That’s why I decided to document the journey — building the resources I wish existed.
Each repository is:
Free and open-source
Fully runnable with uv + Jupyter
Written in clear, modern Python
Designed for clarity, not complexity
It’s an ecosystem — not just of code, but of understanding.
⚙️ What’s Next
The next step in this journey will be 04 — AI Intermediate Projects,
where I’ll take everything built here and turn it into real-world, portfolio-ready applications.
That’s where we’ll explore:
RAG systems at scale
Multi-agent orchestration
Memory graphs and evaluation
It’s going to be the stage where all your foundations evolve into production-grade AI systems.
💬 Final Thoughts
If you’ve ever felt lost trying to connect all the dots between:
Python fundamentals
LangChain and LLM logic
Orchestrators like LangGraph and CrewAI
And building something people can actually use…
Then this series is for you.
Whether you’re a developer, researcher, or curious engineer —
you’ll find here a clear, structured roadmap that takes you from learning to building.
🔗 Explore the Repositories
📦 GitHub Profile → github.com/JaimeLucena
03 — Agents & Apps Foundations
👋 Let’s Connect
If you’re also building with LangChain, LangGraph, or AI agents — I’d love to connect, exchange ideas, and keep improving this learning path together.
📍 GitHub
⭐ If this helps you…
If you find these repositories useful, please consider giving them a star ⭐
It helps others discover this path — and keeps the motivation going to keep expanding it.
Next chapter: Intermediate AI Projects — where things get real.
Stay tuned 🚀
