Timeless

An AI-powered application where you can relive moments, immerse yourself in new realms, re-explore the past, and envision the future.

Comment

Inspiration

Every image tells a story; so does every song.

In the rush of our fast-paced lives, it's easy to lose sight of the present. As students, we become so consumed by our ambitions and goals that we often forget to pause, take in our surroundings, and appreciate the moment. We tend to overlook the beauty of where we are, who we are, and how far we’ve come. When do we ever stop to reflect on life in the midst of our relentless pursuit of success?

That’s why we created Timeless. It’s a space designed for those moments when you want to slow down, revisit forgotten memories, explore feelings of nostalgia, or even imagine the future. As students, we understand the importance of finding balance between ambition and reflection, and Timeless allows you to immerse yourself in those meaningful experiences.

What it does

Timeless is an AI-powered application designed to deliver immersive experiences by generating personalized photos, music, and videos based on user prompts. Leveraging cutting-edge AI models such as Cohere, Groq, DALL-E 3, Luma AI, and Suno AI, Timeless crafts unique audio-visual journeys that capture the essence of specific moments in time. Each experience is enriched with a photo summary, allowing users to fully engage with the atmosphere of the moment. By seamlessly blending creativity, nostalgia, and advanced technology, Timeless redefines how we experience history and memories in the digital age.

How we built it

For the front end, we began by designing each page in Figma, using React.js, Tailwind CSS, and Lucide libraries to enhance the visual aesthetics and overall appeal. We then integrated the front end with the back end by making RESTful API calls and utilizing the Axios package to facilitate smooth communication between both ends.

Our backend consists of Python with the Flask framework, MongoDB Atlas for the database, and Postman for building and testing APIs. We carefully engineered a pipeline among the AI models including Cohere, Groq, DALL-E 3, Luma AI, and Suno AI APIs, ensuring each one generated the input for the next stage in the process. Thanks to Hack the North and their extensive technology stack, we were able to create our backend using these APIs and technologies as efficiently and effectively as possible.

Challenges we ran into

Initially, we had difficulties connecting the front end and back end. After consulting with a few mentors, we realized the best way to integrate them was through APIs and HTTP requests. Developing a simple, functional API was a significant challenge for our team at first.

Timeless relies heavily on AI and requires the integration of numerous APIs. Therefore, prompt engineering and efficient API usage were crucial. Providing each AI model with sufficient information and designing a pipeline to utilize all of them was a time-consuming and difficult task. As a result, our team spent a lot of time fine-tuning prompts, experimenting with different prompt lengths, and testing various conditions to achieve the best results in the shortest possible time.

Accomplishments that we're proud of

  • Our user-friendly and simple interface, carefully designed to make users feel welcomed and relaxed.

  • Our elegant design for pipelining various globally recognized AI models and APIs to create a more powerful and cohesive application.

  • Our application works end-to-end, and we successfully implemented exactly what we envisioned. The final product closely aligns with our initial goals for this hackathon.

  • Our ability to work quickly and stay united as a team, supporting each other and having fun while accomplishing a common task.

What we learned

  • We learned about new APIs that will simplify future projects, especially in the field of artificial intelligence.

  • We developed the ability to work quickly as a team, staying united to complete a shared goal, no matter the challenges.

  • We gained insights into prompt engineering, optimizing AI performance, and reducing time and resource consumption.

What's next for Timeless

Moving forward, we aim to spend more time refining and potentially training our models to make them more customizable and efficient for the project’s purposes. While Timeless introduces a significant innovation, it still needs further development to meet industry standards in terms of quality, design, and security (e.g., reducing long wait times for AI video generation). We plan to implement a more robust authentication system, such as Auth0, which we couldn’t use due to time constraints. Additionally, we aim to introduce new features, like defining events, cultures, and objects for inclusion in users’ final generated videos. We also plan to produce higher-quality videos by using more advanced APIs, potentially beyond Luma AI.

Built With

Share this project:

Updates