Inspiration
The inspiration for Paruni came from my fascination with both celestial themes and the power of natural language processing (NLP) technologies. I wanted to create a unique and immersive experience where users could interact with a cosmic-style website and witness their stories transformed into parallel universes. The concept aimed to blend the beauty of the cosmos with the capabilities of NLP and artificial intelligence.
What it does
Paruni is a website I developed that allows users to input a small paragraph of a story into a textarea. Using NLP techniques powered by the Compromise library and OpenAI's API, the website generates a new story set in a parallel universe while retaining the original characters.
How we built it
To build Paruni, I utilized the Compromise library for natural language processing tasks. Compromise provided powerful features for handling text input and extracting key information from the user's paragraph. I also integrated OpenAI's API to generate the parallel universe story based on the extracted characters and context. The frontend of the website was developed using HTML, CSS, and JavaScript. I focused on creating a cosmic design aesthetic, incorporating animations and visual effects that evoked the celestial theme.
Challenges we ran into
During the development of Paruni, I encountered several challenges. Firstly, finding suitable libraries for the project, particularly for NLP tasks, proved to be difficult. I had to extensively research and evaluate different options before settling on the Compromise library, which provided the necessary features for handling text input and extracting key information. Moreover, creating the animations for the website required careful thought and consideration.
What we learned
Throughout the development of Paruni, I gained valuable insights into the practical implementation of NLP techniques and leveraging external APIs like OpenAI. I deepened my understanding of natural language processing tasks such as extracting key information, handling context, and generating coherent text.
What's next for Paruni
In the future, I'm hoping to improve the capabilities of Paruni by introducing more sophisticated NLP models and techniques. This could involve leveraging advancements in the field, such as contextual embeddings, sentiment analysis, or generating stories with branching narratives.
Log in or sign up for Devpost to join the conversation.