Install uv
For new local AI projects, uv is comfortable because it combines several jobs: package management, virtual environments, and Python versions. That lowers friction when moving from one lesson to another.
curl -LsSf https://astral.sh/uv/install.sh | sh # Windows PowerShell powershell -ExecutionPolicy ByPass -c "irm https://astral.sh/uv/install.ps1 | iex"
Create a clean project
uv init aulafy-practice cd aulafy-practice uv python install 3.12 uv venv source .venv/bin/activate uv add rich requests python-dotenv
On Windows, activate the environment with .venv\Scripts\activate. The important part is that each project has its own environment and declared dependencies.
Minimum Python for AI
- Read and write JSON, Markdown, CSV, and plain text.
- Write small functions with type hints.
- Use try/except for network, file, and model errors.
- Understand basic async when an app calls APIs or slow tools.
import json
from pathlib import Path
def load_json(path: str) -> dict:
return json.loads(Path(path).read_text(encoding="utf-8"))
config = load_json("config.json")
print(config.get("model", "no model"))Aulafy judgment
Start with plain Python. Before adding LangChain, LangGraph, LlamaIndex, or large frameworks, make sure you can make a call, save a trace, and reproduce the result.