Welcome to AI Notebook: Documenting my journey AI
I'm a developer learning to build AI-powered applications, and this blog is where I document that journey. Every post covers something I've built and experimented with for my learning - sharing my learnings for my benefit, but if it helps anyone else, then all the better.
The initial focus will primarily be on LangChain and LangGraph, two of the most widely used frameworks for building LLM-powered applications in Python. If you're on a similar path, I hope this saves you some of the head-scratching I went through.
All code is available in my GitHub repo: ai-notebook
My Setup
Before diving into posts, here's the environment I'm using so you can code along.
Language: Python 3.12
Editor: VSCode (with the Python and Ruff extensions)
Package manager: uv — a fast, modern replacement for pip/venv
If you don't have uv installed:
curl -LsSf https://astral.sh/uv/install.sh | sh
To run any example from the repo:
git clone https://github.com/rob212/ai-notebook.git
cd ai-notebook
uv sync
You'll need LLM API account(s
Most examples use a cloud LLM. You'll need at least one of:
Some examples use Ollama to run models locally for free — I'll flag when that's an option.
Handling API Keys Safely with dotenv
Never hardcode API keys in your source code or commit them to GitHub. I use the python-dotenv pattern instead.
1. Create a .env file (this is gitignored — never committed):
cp .env.example .env
Then open .env and fill in your real keys:
OPENAI_API_KEY="sk-..."
ANTHROPIC_API_KEY="sk-ant-..."
2. The .env.example file is committed to the repo. It shows which variables are needed without exposing real values:
OPENAI_API_KEY="insert_API_Key_Here"
ANTHROPIC_API_KEY="insert_API_Key_Here"
3. Load them in your code:
import os
from dotenv import load_dotenv
load_dotenv()
openai_api_key = os.getenv("OPENAI_API_KEY")
This pattern means you can share your code publicly without ever accidentally leaking credentials.
What's Coming
Posts will follow the structure of my ai-notebook repo, building from the basics up:
LangChain — prompts, models, chains, agents, memory
LangGraph — stateful multi-step agents and workflows
Each post will have a link to the full working code in the repo. See you in the next one.
