Mastering the art of prompt engineering with Empower
It’s 2025, and GenAI is officially mainstream and one of the most transformative
technologies today.
In just two years, it has evolved from a novelty into an indispensable
tool across industries. From generating text, images, and videos to writing code and even composing music,
GenAI is reshaping how we work and create. You might even struggle to think
of the last time you drafted your resume or summarised a report without
it.
That being said, most users barely scratch the surface of its true capabilities.
Why?
Because they’re not asking AI the right questions. In other words, their skills in Prompt Engineering leave much to be desired.
What is a Prompt, and Why Does it Matter?
A prompt is the input (most commonly in text) you provide to an AI model. The way you structure this input directly influences the quality of the response you get.
For example, if your prompt is something as vague and brief as <Tell me about climate change>, it’s unlikely you’ll get output that you might be able to use.
However, if you can feed GenAI with more detail, you’re more likely to get something more useful.
This is where prompt engineering comes in—the practice of designing effective prompts to ensure AI generates accurate, relevant, and high-quality responses.
Introducing the CO-STAR Framework
At GovTech, our officers follow the CO-STAR Framework, featured in our Prompt Engineering Playbook, to craft better prompts.
CO-STAR stands for…
-
Context – Provide background information to help the AI understand the specific scenario.
-
Objective – Clearly define the task to direct the AI’s focus.
-
Style – Specify the desired writing style to align the AI’s response.
-
Tone – Set the tone to ensure the response resonates with the required sentiment.
-
Audience – Identify the intended audience so the AI can tailor the response
accordingly. -
Response – Define the format of the response to get structured, useful outputs.
Applying that to our earlier climate change example, it might look like this:
I work in a non-profit organisation (Context) and I have been asked to write an electronic direct mailer informing people about climate change (Objective). Write this in the style of popular lifestyle publications (Style), keeping the tone informal and friendly (Tone). This will be read by audiences residing in Singapore (Audience) I need this to be about 350 words long, broken up into paragraphs of 100 words each. End with a Call-To-Action to sign up to the newsletter (Response).
That’s just the start, though. The CO-STAR framework is powerful, but
real mastery comes through hands-on practice.
And that’s where the Empower team at GovTech comes in.
Empowering Public Officers
If you think about it, prompt engineering isn’t about engineering—it’s
about applying structured thinking to communicate effectively with AI.
This requires a good understanding of how AI functions.
That’s why the Empower team in GovTech focuses on driving the digital
capabilities of public officers, equipping them with the skills and knowledge
to use AI in their daily work.
One of the key initiatives under this umbrella is the Data & AI Tournament
(DAT) with a prompt engineering award (Prompt Royale) first introduced
in 2023.
In Prompt Royale,
participants are tasked with real-time challenges—everything from generating
text and crafting social media posts to analyzing data sets, building custom
GPTs, and even pitching AI-based solutions—all under the spotlight.
Such initiatives expose officers to the latest tech and help them develop
awareness and interests as well as their capabilities. This way, officers
can leverage AI’s capabilities to solve complex, real-world problems—from
improving public service delivery to streamlining internal processes.
By making prompt engineering a core skill, the team is helping public officers work smarter, not harder. Instead of spending hours figuring things out manually, they can let AI handle repetitive tasks and focus on higher-level problem-solving and innovation.
AI for everyone
If there’s anything that we’ve learned from the Empower Team, it’s that prompt engineering isn’t a fancy term just for people well-versed with AI.
Rather, it’s a practical skill anyone who uses AI—whether for work or personal projects—can tap into to ask better, more important questions and get the results they need.
So, the next time you’re staring at the blinking cursor in the prompt
field, try utilising the CO-STAR framework.
It just might make all the difference in making the AI stars align for
you.
PS: For more on how GovTech is embracing AI, check out GovTech Decoded episodes 5 and 6!
