How We Boosted Productivity in Our Development Process

This is the second post in our Chatbot Chronicles series.

In the last post, I explained how our development team used ChatGPT to turn a client brief into a quick MVP. Today, I’ll show how we use ChatGPT as a pair programmer to speed up development.

What Is Pair Programming?

Pair programming lets two developers work together on the same code. It boosts code quality, reduces bugs, and promotes knowledge sharing. Our team at Sauce often uses this approach, and while it can feel slower, research proves otherwise. Even so, finding time to pair with a teammate isn’t always easy.

The most common style is Driver-Navigator:

  • The navigator checks logic, suggests improvements, and thinks ahead.
  • The driver types and navigates the codebase.

Robot Navigate! Using ChatGPT as a Pair

We started experimenting with ChatGPT as a permanent navigator—always ready to help. Developers use it to fix bugs, review documentation, suggest refactors, or parse data.

Here’s how it works:
A developer hits a problem, asks ChatGPT a question, and provides the code. ChatGPT reviews the snippet and responds with ideas or cleaner solutions. The developer then applies or modifies the suggestions.

In one session, I asked ChatGPT to refactor some React code into a reusable Button component.
Knowing I’d want color control, I requested a version with custom styles.

ChatGPT’s first solution had issues, but after I gave feedback, the second attempt hit the mark. The back-and-forth mimicked real pair programming. The interaction felt natural—like coding with a teammate.

A refactored example of a button as a react component

Following this thread, I know that I’m going to want to control the color of this button. Let’s ask the bot to have a go.

An example of how you can modify the code to achieve ai pair programming productivity

Why It Works for Us

ChatGPT saves us time. Instead of booking a teammate for a session, we tap into AI when needed.
It offers a second opinion and sometimes suggests ideas we wouldn’t consider.
It also cuts down on repetitive typing—a welcome relief for anyone with RSI.

Of course, it’s not perfect.
It doesn’t fully grasp project-specific context or catch every bug.
But as a supplement to human collaboration, it boosts our productivity and enhances how we work.