Eaze Read | Devpost

A conceptual idea of what our project does. From articles filled with nonsense to neat summary to bionic design.

Inspiration

We thought it would be cool to build a Chrome extension, as it isn't challenging to make one and we already had many necessary skills. For example, a "hello world" Chrome extension can be created in just a few minutes with some HTML/CSS/JS code. We wanted to create a project like this, but with functionalities that could make it easier to read from an article.

What it does

With Accessibility in mind, Our project has two main features.

Making summaries about articles and displaying them in bionic format.
Making complex text-based articles, easier to read and understandable. It will be very useful for children and elderly people to stay up-to-date with trends and NEWS.

How we built it

Eaze Read, as the name suggests, makes content easier to read. The technique uses suggestive AI in our brains. Our brain has already seen most of the words in the English/Latin language. It can recognize the word by reading just the first few characters of it and the context under which it was used. We'll just make the beginning characters bold, depending on the size of the word.

We have used NodeJS (JavaScript framework) for the backend. This backend acts as a middleware for our Chrome extension and the OpenAI API server. The Chrome extension is built using a basic tech stack. It's HTML, CSS, and JavaScript. Other than that, We have deployed our backend on Vercel.

Challenges we ran into

All the summary generative AI tools are paid. So it was a tad difficult to get a hold of the API for OpenAI's GPT 3.5 Turbo model. Communication between different modules took most of our time. As you cannot just directly access the site content from the Chrome extension and vice versa.