Getting Started

Python Setup

Setting Up

There are two starting points for this programme, and which one you need depends on what you are doing today.

  • Google Colab — for learning Python. Nothing to install, works in a browser.
  • Antigravity — for the hands-on sessions, where you build real agents on your own machine.

Start with Colab. Move to Antigravity before your first hands-on session. If you have never run Python at all, do Your First 30 Minutes first.

Google Colab

Go to colab.research.google.com, sign in with a Google account, and click New notebook.

That is the entire setup.

  • nothing to install, nothing to configure
  • identical on Windows and Mac
  • saves to your Google Drive automatically
  • free

Installing Python on a work laptop is where a lot of people give up on day one — version numbers, PATH problems, permissions, a machine that will not let you. Skip all of it while you are learning the language.

Rename your notebook Python Practice and use it for the whole of this tutorial.

Antigravity

For building agents you need a real editor on your own machine. This programme uses Antigravity.

Why this one:

  • it is a vibe-coding editor — you describe what you want and it writes the code
  • it is by Google
  • it is free, within limits that are comfortable for what we do

That last point is the deciding one. Plenty of excellent tools exist — Claude Code, Copilot, Cursor, Kiro — and many people already have licences through their employer. But nobody should have to buy a licence to take part in this programme, so the default is the one that costs nothing.

If you already pay for something else and prefer it, use it. Nothing here depends on the editor.

Setup instructions: link.futureproofindia.com/setup

Follow that page rather than improvising. It has the installation steps and the prerequisites for the hands-on sessions, and those prerequisites are not optional — turning up without them means spending the session installing instead of building.

Use your own laptop, not the company one

Worth saying plainly, because it catches people every cycle.

A corporate laptop often blocks software installation and restricts outbound API calls. Both are things you need. If you have a personal laptop or PC, use it for this programme. It will save you a great deal of frustration.

Installing Python locally

You do not need this for Colab, and you may not need it at all if Antigravity manages it for you. When you do:

python --version

If that prints 3.11 or higher, you are ready. On some machines the command is python3; on Windows, py --version is often the most reliable.

Otherwise download the current release from python.org.

On Windows, tick "Add Python to PATH" on the first screen of the installer. It is easy to miss and it causes most "python is not recognised" problems afterwards.

Virtual environments

This is the part people skip, and the one that matters.

A virtual environment is a private folder of packages belonging to one project. Without one, every install goes into a single shared pile and two projects needing different versions of the same library break each other. Java and C# isolate dependencies per project automatically. Python does not — you have to ask.

python -m venv .venv

Activate it. On Windows:

.venv\Scripts\activate

On macOS or Linux:

source .venv/bin/activate

Your prompt gains a (.venv) prefix. That prefix is how you know it worked.

Installing packages

python -m pip install httpx python-dotenv

Use python -m pip, not a bare pip. It guarantees the install goes to the interpreter you are actually running, which removes the most confusing category of Python problem — a package that is installed but "not found".

Record what you installed:

pip freeze > requirements.txt

And restore it elsewhere:

pip install -r requirements.txt

A .gitignore, from day one

.venv/
__pycache__/
.env
*.pyc

The .env line is the one that stops you committing an API key to a public repository. It matters more than everything else on this page.

When to leave Colab

Colab is genuinely good, and plenty of useful work never leaves it. Move to a local setup when you need:

  • more than one file
  • secrets not tied to your Google account
  • version control
  • something that runs on a schedule or serves requests
  • a long job a browser tab would drop

Do not move before you need to.

Common problems

SymptomCause and fix
python is not recognisednot on PATH — re-run the installer and tick the box
ModuleNotFoundError after installingwrong interpreter — use python -m pip, check sys.executable
Permission errors on installinstalling system-wide — use a virtual environment
Cannot install anythingcorporate laptop restrictions — use a personal machine
Works for you, not a colleagueyou never wrote requirements.txt

Practise this

  • Open Colab, create a notebook called Python Practice, and run print("Hello World").
  • Follow link.futureproofindia.com/setup and get Antigravity installed before your first hands-on session.
  • Create a virtual environment, activate it, and confirm the (.venv) prefix appears.
  • Write a .gitignore containing .env and .venv/.

Try It Yourself

Copy this into a new Colab notebook and run it. Nothing to install.

example.py
1# Check Python version
2import sys
3print(f"Python version: {sys.version}")
4
5# Check if common AI packages are available
6packages = ["httpx", "python-dotenv", "pydantic"]
7
8for pkg in packages:
9    try:
10        __import__(pkg)
11        print(f"✓ {pkg} is installed")
12    except ImportError:
13        print(f"✗ {pkg} is not installed")