Python Essentials

🐍 Python Essentials (goal: productive for scripting + AI)

You don’t need Python mastery. You need enough to use the AI ecosystem (LangGraph, evals, notebooks), where Python is the first-class citizen.

Checklist

• Syntax, comprehensions, dataclasses, type hints, f-strings
• Virtual envs + package management with uv (modern and fast); pyproject.toml
• asyncio basics (async/await, gather), since AI SDKs are async-heavy
• Pydantic (data validation), used everywhere in AI tooling
• FastAPI basics (to expose a Python agent as a service)
• Jupyter notebooks for experiments
• requests/httpx, pytest

🧪 Labs (🟢 warm-up → 🟡 core → 🔴 hard → ⚫ boss)

• 🟢 A uv project + pytest + ruff
• 🟡 FastAPI service with Pydantic models + SSE streaming
• 🟡 asyncio: fan out 50 HTTP calls with a semaphore; compare with Go

🧠 Cognitive tasks

• Transfer: map Python async concepts to Go goroutines and Java virtual threads

🛰️ Orbit integration

• orbit-agents (LangGraph) and orbit-evals are Python services

Go deeper

🧩 AI & Agent Patterns

Resources

• Official Python tutorial · Fluent Python 2e (reference for later) · FastAPI docs