Inspiration
We wanted to build an immersive, interactive experience that pulls users into a dystopian, steampunk world. Drawing inspiration from classic adventure stories and role-playing games, Tail of Time lets users journey through a unique, action-filled storyline where they interact with a quirky crow and navigate an adventure in a constantly evolving timeline.
What it does
Tail of Time is an AI-driven storytelling platform where the user, playing as a simple old man, is guided by a clever, humor-filled crow through a series of steampunk-themed adventures. With choices presented at each turn, users can shape the path of their journey, unlocking new scenes, characters, and events as they progress. The AI dynamically generates responses, keeps the story moving, and makes sure to offer choices—like “A, B, C, or D” options or open-ended questions—creating an interactive, entertaining experience filled with action, humor, and surprises.
How we built it
We built Tail of Time using Cloudflare Workers to create a seamless backend for generating AI responses in real-time. The AI model was configured to retain memory of previous interactions, giving a sense of continuity and allowing the story to progress naturally with each choice the user makes. Each interaction dynamically adjusts to the user's choices, ensuring that the story feels personalized and unique every time.
Challenges we ran into
Getting cookies to work was one of our primary challenges, as we needed them to persist across sessions to maintain the continuity of the storyline. The cookies initially kept refreshing with each page load, causing the user’s progress to reset unexpectedly. Configuring Cloudflare was also a learning curve since it was our first time using the platform for this kind of AI-driven project. Setting up Workers, managing environments, and handling cross-origin requests took trial and error, but ultimately, it allowed us to create a responsive backend that communicates well with the frontend.
Accomplishments that we're proud of
We’re proud of creating a dynamic, story-driven AI experience that feels both responsive and engaging. The AI’s ability to remember past interactions and advance the story gives users a continuous, immersive experience. We’re also thrilled with how well the character-switching feature works, allowing users to interact with multiple perspectives and personalities within the storyline, which makes each interaction fresh and exciting.
What we learned
We learned a lot about cookie management, prompt engineering, and Cloudflare’s capabilities. Working with cookies to retain user-specific data was crucial in creating a smooth, consistent experience. Fine-tuning the prompts to guide the AI effectively while ensuring it followed the storyline took some experimentation. Setting up and optimizing Cloudflare Workers taught us the importance of effective backend configuration to keep our AI interactions efficient and fast.
What's next for Tail of Time
Our next goal is to make the AI even more interactive by teaching it to trigger specific animations in response to story events. For example, if the AI mentions “bats,” it could trigger an animation of bats flying across the screen, making the experience visually engaging and adding a new layer of immersion. We also plan to enhance character-switching to make the AI transition seamlessly between roles, adding depth to the storytelling. Additionally, we want to introduce diverse background art to enrich the steampunk aesthetic and make the user interface visually captivating, helping users feel even more deeply immersed in the world of Tail of Time and with maybe with a TTS (Text to Speech) so user does not have to read to long passages.
Log in or sign up for Devpost to join the conversation.