How Our Team Built a Video Game Using Generative AI Assistant Amazon Q Developer

"The gravity needs to go in the opposite direction," Mission’s graphic designer Krista Clarke explained to the generative AI chatbot. Two days into developing her first game, she was already solving complex programming problems through conversation. Meanwhile, our CTO Jonathan LaCour discovered that Amazon Q Developer, the generative AI assistant he used, could transform his coding workflow in unexpected ways.

At AWS re:Invent 2024, visitors to Mission's booth #1954 could play Krista’s and Jonathan’s creation — a game called Skyfall where you guide a skydiver through clouds while dodging helicopters, competing for a PS5 Pro. But the real story is about how this game came to be and how generative AI is changing the software development landscape.

The Challenge: Creating an Interactive Game Experience

When planning our AWS re:Invent booth strategy for 2024, we wanted something more engaging than a typical product demo. We decided to build a game that would be both fun and interactive while demonstrating the power of AI.

Instead of hiring game developers or outsourcing, Krista, our Graphic Designer, who knew basic HTML, paired up with Jonathan, our experienced CTO. The goal was straightforward: make a game where players navigate a skydiver through an obstacle course, collecting points from clouds while avoiding hazards.

Building this type of game typically requires deep knowledge of programming languages and game development frameworks. At least, that's how it used to be.

How a Designer and a Developer Leveraged Gen AI to Build the Same Game

Starting from Scratch: The Designer's Story

"I've always been something of a gamer," Krista explains, "but I've never done any sort of game development before." With just self-taught HTML and limited JavaScript knowledge, she turned to generative AI to bridge the knowledge gap.

What would typically take weeks of learning programming fundamentals became an intuitive conversation with a generative AI chatbot. Krista found that her gaming experience was more valuable than coding knowledge: "More than development experience, you just kind of have to have a knowledge of games in general. You have to know what to ask for."

Within just two days, she had created a working prototype using Python and the Pygame framework — technologies she hadn’t had experience with before. The AI helped her implement everything from character movement to collision physics, scoring systems, and basic game mechanics. When technical challenges arose, like gravity flowing in the wrong direction, she simply explained the desired behavior to the AI assistant: "You have to be able to talk about physics and collisions, and more just kind of game speak as opposed to necessarily the development behind it."

Even more impressively, when the need arose to convert the game from Python to JavaScript for web deployment, Krista worked with AI to transform the entire codebase. "After having my game running successfully in Python, I had to leverage AI to convert the whole thing to JavaScript, which was a whole new adventure," she recalls. The AI guided her through hosting solutions and helped her create a testable prototype on our website.

Taking it Further: The Developer's Perspective

Despite 30 years of coding experience across many different projects, Jonathan had never developed a game before this project. When he took over from Krista to enhance the game further, he discovered Amazon Q Developer's transformative impact on development workflows. "It would have taken me weeks instead of hours," he reflects. "Within about four hours, I had the first prototype completely functional."

But what surprised him most wasn't the speed — it was how Amazon Q Developer, the gen AI assistant he used in this project, transformed his development process. "My typical workflow when I want to write some code is to start with some form of basic research... It's an exercise for me to find all the information out there. This was a very different experience."

Amazon Q Developer served as an intelligent pair programmer, helping Jonathan focus on architecture and design decisions rather than getting bogged down in implementation details. "I find myself not needing to go to the internet or to refer to documentation hardly at all. Instead, I am asking questions and getting back code that is tailored to me and my question."

Breaking Down the Development Process

The game evolved through constant dialogue between humans and AI. Rather than spending hours searching documentation and Stack Overflow, both developers could focus on creative problem-solving. For Krista, this meant iterating on game mechanics and user experience. For Jonathan, it involved architectural decisions and performance optimization.

The development process became more intuitive and conversational. Instead of hunting through documentation to implement a feature, Krista and Jonathan could simply describe what they wanted to achieve. Amazon Q Developer would suggest implementation approaches, generate code snippets, and help troubleshoot issues — all while explaining its reasoning in plain language.

The Power of AI as a Development Partner

"People often worry that AI will write buggy or unreliable code," Jonathan explains. "But that misses the point. AI is a development tool, not a replacement for developer judgment. Yes, the code it generates might have bugs — just like any first draft of manually written code. My role is to review, test, and refine that code. AI helps me write code faster, but I'm still responsible for ensuring it works correctly."

This partnership approach to development suggests a future where AI doesn't replace developers but amplifies their capabilities. As Jonathan notes, "Does that mean that we're going to need fewer programmers, or does that mean we're going to have programmers who are 10 times as efficient as they were before? I tend to think the latter is the case. There's no shortage of problems in the world to solve." Every business has miles-long backlogs for their software. Everyone wants to build a lot of things simultaneously, and we don't have the capacity to do it right now, which is why AI is so powerful in software development.

The project demonstrated that generative AI can adapt its assistance based on the developer's experience level. For Krista, it provided more comprehensive explanations and handled more of the technical implementation. For Jonathan, it served as a sophisticated coding partner, helping him move faster while maintaining control over the architecture and code quality.

A Glimpse into the Future of Software Development

This experiment revealed something profound about the future of software development. When a designer and a CTO can use the same tool to build functioning software, we're witnessing the democratization of development. The traditional barriers between technical and non-technical roles are starting to fade, creating new possibilities for innovation.

"It all just felt very natural to me," Krista reflects. "With generative AI, I never really felt like I hit a wall with anything." Her experience suggests that the future of software development might be less about memorizing syntax and more about clearly communicating ideas and understanding core concepts.

For experienced developers, AI tools like Amazon Q Developer aren't just about writing code faster — they're changing how developers approach problems entirely. Rather than spending time searching documentation or Stack Overflow, developers can focus on architecture, design, and creative problem-solving.

The success of this project demonstrates that AI is more than just another development tool — it's a bridge between vision and execution, making software development more accessible while helping experienced developers work more efficiently.

Didn’t make it to AWS re:Invent this year? No problem. You can play the game directly on our website. If you want to explore the code and start building a game yourself, check out Jonathan’s open-source implementation on GitHub.