It's always been a struggle to produce coherent, professional, and compelling writing. When I came across the MeaningCloud API, I thought about how useful it would be for editing essays, as a student. And it became an even more important idea when I realized that it could be very powerful for somebody in marketing or politics to edit their speeches.

What it does

Analyzes your document and provides insights about what you're writing. It pops out a sidebar with information about your document, like if it is objective or subjective, if it is ironic, if it is coherent, how you can make it more coherent, and more.

How I built it

I used Apps Scripts and G Suite SDK to read the document and create the sidebar. I utilized the MeaningCloud API to run various analyses on the document, and then used the web stack to create the webpage (the graph is build with Chart.js).

Challenges I ran into

It was very difficult to understand the G Suite Marketplace SDK at first, because it has is a new framework, and requires you to work in a special environment you write your code. Additionally, it wasn't easy to construct the API request for MeaningCloud, but once these two obstacles were passed, it was smooth sailing.

Accomplishments that I'm proud of

I'm elated that I was able to build this app well, and submit it for publication to the the Chrome Web store and the G Suite Marketplace. It looks like it belongs in Google Docs, and I've had many friends sign up for testing it out.

What I learned

I learned about integrating applications together, and how challenging it can be to have the mental overhead of three different architectures at the same time. It was very challenging, but once I put together a framework that handled these three topics well, it was (again) smooth sailing.

What's next for Concept

I would love for Concept to become real-time, which isn't the case right now because I don't want to run out of MeaningCloud API credits. Additionally, more insights, and better coherence detection would be a major goal.

Built With

Share this project: