AWS Logo
Menu
Ditch Your Diary & Play Your Feelings: Meet Chaperone

Ditch Your Diary & Play Your Feelings: Meet Chaperone

Ever feel your life is a story too big for words? With Chaperone, you turn your experiences and wildest ideas into immersive, story-driven games.

Published Jan 12, 2025
Last Modified Jan 14, 2025
Ever wished you could share your life experiences in a deep, impactful, and interactive way? Or maybe explore someone else's story and imagination meaningfully? That's exactly what my teammate Yuhang and I set out to create with Chaperone! Chaperone is all about the power of storytelling – the idea that anyone can create and share interactive stories. We're talking Choose-Your-Own-Adventure style narratives but with a serious upgrade. You can dive into real-life experiences, explore fantastical worlds, or vent about that awful day at work, all through the medium of player-driven narrative games.
Here's the breakdown:
  • Play: Explore a growing library of story-based games created by other users. Each choice you make leads to a different outcome, giving you the possibility of a unique adventure every time.
  • Create: Got a wild dream? A funny memory? A pet peeve about pineapple on pizza? Turn it into a game! Just input a detailed prompt, and our AI gets to work creating a game you can share with friends and family.
  • Connect: Favorite games, build a personal library, and see what's trending in the community. It's all about sharing experiences and connecting with others on a deeper level.
  • Immerse: Control music and effect levels for an immersive experience.
Sneak Peek
Each game you create has 16 distinct endings for every story thanks to the way we structured our algorithms, this leads to dramatic twists and turns every time a game is played depending on the user's choices. I used the Fibonacci sequence to simulate an insane exponential growth in the possibilities of outcomes in the game as the game story progresses and this increases the replay value. You basically never have the same play-through the first few times you play a game depending on the choices you make.
We wanted to make it easy for anyone to communicate and share their stories or feelings with the world through the medium of gaming and storytelling, and we think this is a cool way to do it!
Amazon Q is the star of this show and helped me complete the dev work on time for the AWS Game Builder Challenge despite joining the hackathon a bit late. Can't wait to check it out?

The AWS Dream Team: Q, S3, and Bedrock

Here's where the real AWS magic happens:
  • Amazon Q: My Code Whisperer: Amazon Q is a game-changer. This wasn't just a code completion tool – there were moments where it felt like it was reading my mind! The context awareness was impressive. I’d write code in one file, and then Q would make suggestions in a different file related to the first, like it knew exactly what I wanted. I'm not gonna lie, it got a little eerie at times, but in the best way possible. Plus the ease of switching between inline and chat mode in the development process was very convenient. It sped up the dev process SIGNIFICANTLY. I mean, we got all the core functionality to get it ready to be published on three platforms (Android, iOS, and web!) within a space of 2 weeks during the Christmas and New Year holidays - and Amazon Q played a major role in making that feasible.
  • Amazon S3: Our Digital Vault: All those game assets – from character art to UI elements – live in S3. Setting up S3 was surprisingly smooth, especially the CORS and bucket policy JSON configuration. It's refreshing to see that a good blend of security and ease of use is possible. We anticipate that user-generated content will grow exponentially when we make game asset generation accessible within the game, and S3 with its scalability seems like the logical choice. I've used other cloud storage providers in the past but the ease of use I experienced while setting up s3 got me hooked.
  • Amazon Bedrock (Image Gen): We also dabbled with Amazon Bedrock's Image Gen (Titan Image Generator G1 v2 and Stability AI) for some initial game assets. The results were pretty impressive, and it got me hyped for using it internally within the app for image asset generation in future iterations.

Development: A Rollercoaster of Code, Coffee, and AI

Our development journey was a beautiful blend of excitement and panic. Yuhang did a fantastic job on the UI/UX design, creating an intuitive and engaging experience, while I dove deep into the code and AI integration drama. The biggest challenge? Managing that exponential growth in possibilities with a Fibonacci sequence-based game creation algorithm. It was a tedious logic puzzle! Keeping it all coherent was like herding cats with laser pointers – lots of unexpected jumps and zigzags. There were definitely a few moments of “Why did I think this was a good idea?”.
  • Lessons Learned: Game development is different from regular software development. There is a different mindset required to be a good game dev, and this was a beautiful experience tackling those shifts and challenges. Also, AWS makes its SDK integration a breeze with top-notch documentation.
  • Distance is no Barrier: Despite living in Nigeria while Yuhang is based in Germany, collaborating across borders was seamless, thanks to our similar time zones, which made scheduling development meetups much easier. Additionally, our diverse cultural backgrounds brought unique perspectives and enriched the creative process, making Chaperone an even more dynamic and robust game.
  • Amazon Q's Impact: Let me just reiterate, Q is a cheat code. It wasn't just a suggestion box; it was a collaborator. It helped me iterate faster, spot errors quicker and basically kept me from losing my mind on more than one occasion.

User-Made Game Previews

Game Previews

What's Next?

Game text generation: We're looking to dive deeper into Amazon Bedrock, using it to compare the strengths, weaknesses, and biases of different AI models and deploy it for creating our game text content. We're hoping that Amazon bedrock will enable us to combine the strengths of different AI models to create even more compelling games in future iterations of Chaperone.
Game asset generation: Currently, game assets are manually generated but we are also working towards using Bedrock to give users the option of generating image and video assets for their games from within the app. I was particularly impressed by the image results on Titan Image Generator G1 v2.
Titan Image Generator G1 v2
Titan Image Generator G1 v2 Generated Image
Prompt: A person and their dog, both looking content and loving, sitting together in a peaceful setting.
Game Asset Handling: User-generated assets need a comfy home so programmatically storing user-generated assets on S3 seems like a very logical choice.
Backend database manipulation: We're also planning on using Serverless Lambda functions to handle more of our backend workflows as the game grows in scale and complexity.

Your Turn: Share the Story

We're super excited for you to check out Chaperone! It's available on the Play and App Stores (Yup! Despite starting work on this project in late December we were able to get the first release finished and published on time on three major platforms thanks to Amazon Q helping to 10x the development process 🥳):
Give us some feedback! Share your thoughts, and let us know what you think about Chaperone!

A bit more on the tech

  • Amazon Q
  • Amazon S3
  • Amazon Bedrock (Titan Image Generator G1 v2)
  • Vertex AI
  • Llama 3
  • Node.js
  • Flutter.dart
  • Firebase

Final Thoughts

This is my first time making a game also my first experience working with AWS, and I gotta say, it's been awesome. The documentation is top-notch, and the community is incredibly helpful. I'm hooked! For my next project, I'm definitely going all in on AWS. I'm curious to see how far this rabbit hole goes.
Keep The Games Coming,
Increase
 

Comments