databugbattle.io

Made by: Eucharist Tan, Vincent Jones, Kaden Chan

Inspiration

We were inspired by an existing game called generals.io, which is a competitive online multiplayer game with a simple concept to increment numbers and take over opponents. Admiring the simplicity and concept of strategically building links between titles to overtake other players, we decided to create a similar game with this concept of connectivity to illustrate how data centers operate and the impact of a major disruption on an overall network.

What it does

Our game serves as an educational tool for new engineers by simulating how data centers form as a system, the importance of how multiple data centers increase resilience to prevent overall disruption of a network and reduce single point of failures.

How we built it

We used an The Elephant-Goldfish Model (EGM) to design databugbattle.io.

First, we brainstorm a list of ideas that could benefit data center teams. We chose to create a game because we wanted to innovate a product that could help give awareness to engineers on how the severity of a data center disruption could impact the availability, accessibility, and performance of a system. Through simulating how datacenters work together, we hope to provide a hands-on experience to understanding connectivity and resiliency. After establishing an idea, we documented a list of acceptance criteria (the elephant) that we wanted to define our product such as considering our layout of the game as a 20x20 grid, defining what a turn consists of, and how to resolve datacenter outages. We prompted our idea using ai and used models to refine and edit to our desired specifications and preferences into a design document. That design document is then tested with another prompt (the goldfish), which is then refined into our final game.

Challenges we ran into

Establishing a baseline for our game found to be difficult. We found it challenging to communicate thoughts when designing a new game from scratch. What we found helpful was collaborating notes on a google document and handdrawing/bulleting notes on paper, which helped us formulate well rounded ideas. Additionally, brainstorming game mechanics to make the game more interesting was challenging as we were afraid our baseline idea would not be appealing or enjoyable. However, we found that it was helpful to prompt ai to challenge our existing ideas. This helped flesh out unformed ideas and helped define preconceptions that was necessary to making our game successful. While it felt tedious as we took multiple tries to prompt feedback and allow for team discussion, we felt more confident the more we did so.

Accomplishments that we're proud of

We are proud of utilizing ai tools to help create a new game entirely from scratch based on an effective prompting technique. It was our first time using the Elephant-Goldfish Model (EGM), and it really surprised us on how we could utilize it and make our prompting more effective.

What we learned

We took away strategies on effective ways to prompt, how to use ai models to refine ideas, and how to layer models on top of each other to build something unique. We find it impressive how ai models can work together to boost efficiency and create ideas into life in a way that has dramatically changed the way we make software today. This provides opportunities for many with little experience with code to innovate well designed and fleshed out applications.

What's next for databugbattle.io

Hosting the game as a multiplayer game with different gameplay modes (such as competitive, solo, and team multiplayer)! We think that creating different variations of this simulation could be really fun for all people, even those who are not engineers. You could be a rising startup, fighting against the AWS data empire lol.