Back to all modules
MODULE 05

Automate Your Pipeline with Codex

What it is

Codex is OpenAI's AI software engineer. You don't need to code to benefit from it — you (or someone on your team) describe what you want in plain English, and Codex writes, tests, and can even run the code.

The goal is for you to (a) recognize when a task has crossed from "a GPT can do this in chat" into "this needs a real automation/script," and (b) be able to describe that need clearly enough that Codex (or a developer using Codex) can build it fast.

Why it matters

When a recurring task is too complex for a Custom GPT alone — like pulling live data from an API — you now know the phrase to use: "Can we build this as an automation?" rather than doing it manually every week.

Video resources

Video 1: OpenAI Codex Demo — A demonstration of Codex turning plain-English requests into working code.

Workshop

Live Demo

The facilitator uses Codex live to build one small, real automation end-to-end from a plain-English request, e.g.:

"Write a script that reads a Zoho CRM deals CSV export, cross-references it with our target account list, identifies prospects we haven't contacted in 30+ days, and generates a prioritized re-engagement list with a personalized subject line for each."

Watch Codex write the script, run it against a sanitized CRM export, and produce the re-engagement list — emphasizing this took minutes, not a dev sprint.

Hands-on exercise

No formal hands-on in this module

This module requires more setup than is available in-session.

Optional homework

Get a Codex/ChatGPT Business seat with Codex access and try rebuilding your Module 4 Sales & Marketing Automation Blueprint as a real script with facilitator support in a follow-up session.

← Back to all modules