Prompt Engineering for Test Design: Supercharge Your AI Skills

Learn Prompt Engineering techniques to supercharge your GenAI skills and create effective test designs. This post provides a user-focused overview of the GenAI landscape and practical prompting tips

GenAI landscape

Learn the GenAI landscape from a user perspective and prompting tips to generate a test design for the Google Home page

Overview

It’s 2025, and at this point, GenAI does not need an introduction. November 30, 2022, shook the world when Open AI launched ChatGPT and captured the imagination of people worldwide.

We live in a world now where we can ask GenAI to solve different kinds of problems for us like

1. Generate code

2. Edit multiple files

3. Generate images, videos

4. Ask questions about problems and incrementally build solutions

It’s fascinating to see this technological progress, even though there seems to be a lot of noise and hype around it.

Many people are putting out clear signals. I feel like we all got a cool new technology upgrade added to our toolchains as a professional software engineer, and we should leverage it in the right way

I’ve so far been silently watching this space evolve. I use many of the tools and techniques in my daily workflows, and I feel it's time to put my take on how AI is helpful for testing.

I’m by no means an expert but I’ll build and learn in public and hopefully take you with me on this learning journey.

My current mental model 🧠

Let’s adopt the POV (point of view) of a senior engineer focussed more on the user layer and how can they leverage and integrate GenAI in their day-to-day workflows

Side note: Stephen Wolfram wrote a fascinating piece on What Is ChatGPT Doing … and Why Does It Work? I recommend reading this if you want a better intuition on how this technology really works under the hood.

Interfaces 📱

I see 5 main ways to interact with GenAI and get it to do something helpful for you

1. Chat: Type a question in a chat box and see a streaming answer being typed back

2. Inline chat: While coding you can type a prompt in the context of the code you are working on

3. Voice: Speak your prompt out loud in natural language and hear the response back

4. Autocomplete while coding: Coding assistants predict what you want to write and then provide an autocomplete suggestion

5. Multi-file editing: Coding assistants provide context on a few files where you want to make a change and then type a prompt for GenAI to make that change for you.

Input modes ➡️

We can write prompts or type in your question or task along with the required context such as:

1. Ask a question and follow up

2. Prompt to make an edit and accept/discard

Outputs ⬅️

And get back a result in the below formats that you can then either use as is or refine further in more incremental cycles of back and forths

1. Text

2. Code generation

3. Transform text into images or videos

Depending on your use case you may use one or more of these interfaces to get to the desired outcomes.