Rome | Devpost

Inspiration

Imagine a world where brands aren’t just seen but truly felt—where every interaction is an immersive experience that leaves a lasting impression. With Rome, we're crafting that future. We're giving creators and companies the tools to transcend traditional boundaries by turning their unique brand identity into something people can play, explore, and connect with. Every game becomes a reflection of your values, an extension of your aesthetic, and a bridge to deeper engagement with your audience. Through AI-driven personalization, we're not just building games; we’re building relationships, experiences, and loyalty that resonate on a whole new level.

What it does

With Rome, we're giving brands, influencers, and companies the freedom to shape their digital presence like never before. Our application harnesses the power of generative AI to interpret the unique essence of each brand and transform it into a dynamic, personalized video game. Whether you're looking to create an engaging experience for your audience or expand your brand’s digital footprint, Rome delivers a customized gaming solution tailored to your specific needs. Through interactive gameplay, we help you not only stand out but also build stronger connections with your audience, making your brand more memorable and impactful in the digital age.

How we built it

We built the front end of our project using React, TypeScript, JavaScript, HTML, and CSS. For the back end, we relied on Django and several Python libraries to handle image manipulation. We also utilized an unofficial Midjourney API, also written in Python. To create the games, we used Kaboom.js, along with HTML and CSS.

Challenges we ran into

Throughout the development process, we faced numerous challenges. The Midjourney API was particularly tricky, as there's no official version, and the unofficial ones essentially prompt Midjourney in a Discord server and then fetch the generated images using Discord API keys. Another challenge with using Midjourney or any AI in general is that they are non-deterministic by design. This leads to challenges in things like having the original generated images be suitable for being converted into sprites or being able to get multiple of the same sprite for in game animations. Connecting the front end and back end also presented its own hurdles. Additionally, setting up and getting the APIs to work for game development was no easy feat.

Accomplishments that we're proud of

Despite these obstacles, we're extremely proud to have completed the project within a tight 36-hour timeframe.

What we learned

We learned a tremendous amount throughout this journey, gaining proficiency in React, TypeScript, JavaScript, HTML, CSS, Django, various Python image manipulation libraries, and the inner workings of an unofficial Midjourney API.

What's next for Rome

Moving forward, we plan to beta test "Rome" and enable the generation of even more assets and sprites. We also aim to enhance the animation and overall enjoyment of the games, potentially even transitioning them to 3D.