Mangalator

• 

  Original Raw

• 

  An example translation

• 

  Original Raw

• 

  Another example translation

• 

  Example translations

• 

  Example translations

Inspiration

What it does

Mangalator is a unique platform designed to empower mangakas by helping them reach a wider audience. By allowing artists to upload their work, which is then automatically translated using our advanced DeepL translation system, we make their creations accessible to readers worldwide. Our primary goal is to help mangakas overcome the language barrier that often hinders their work from gaining the popularity it deserves. With Mangalator, we're bringing down these barriers one manga at a time.

How we built it

Mangalator was built using a modern tech stack to ensure optimal performance and user experience. We used React.js for the frontend to create a dynamic and responsive user interface. The backend was developed using FastAPI, a high-performance framework for building APIs. We chose MongoDB for our database due to its flexibility and scalability. And lastly, we integrated DeepL for the translation functionality, which is at the heart of Mangalator.

Challenges we ran into

We ran into some challenges with deciding on how to represent images under the hood. The original plan was to write them to PNG files, but we found that dynamic imports presented too great of a challenge. We finally opted to pass them as base64 strings.

Accomplishments that we're proud of

The model is surprisingly adept at providing natural translations, taking into account certain nuances. We genuinely think that this provides a strong starting point for mangakas to show their works to a wider audience.

What we learned

We learned that bad documentation has severe impacts on mental health. Modulation works very well, everyone has their own domain to work on, and we were able to multithread and threads.join() at the end. Setting up sucks. Communication is important, gotta plan everything out, just like the thread scheduler!

What's next for Mangalator

Hopefully deployment, and we hope to see this actually be used in real life. Also improve the authentication so it is more safe and robust. We also want to improve the task scheduling on the server side, so it scales better with more users.

Built With

• deepl
• fastapi
• mongodb
• react