Smart Summary

Inspiration

Simply put, Smart Summary saves Business teams time by providing a synopsis of any piece of information, all with the power of robots!

Many world class businesses depend on Confluence to organize internal knowledge. Although well-written, detailed articles are important references for team members, sometimes the team member just needs to know the most pertinent information that relates to the task at hand. Smart Summary improves comprehension and saves the user time by compiling a summary of any Confluence page through the use of artificial intelligence and natural language processing.

What it does

Smart Summary leverages the Expert.ai Hybrid Natural Language Platform to compile a summary of any page or blog post in Confluence, along with an audio file version of this summary, and information on key terms and topics.

More than just merely being a "cheat sheet", Smart Summary can even be used after reading the full length of a document in order to boost comprehension and retention.

Also, taking advantage of Expert.ai's capabilities, Smart Summary includes a "Sentiment" tab that will give insight into the general tone of the article, which can help editing and revising internal content.

How we built it

Smart Summary is a Forge app developed for Confluence. It uses two APIs:

1) expert.AI - This API uses the power of artificial intelligence to extract key takeaways and phrases from each document. This is used to compile a summary that aims to provide key information from each page, saving the user time and increasing efficiency. Key topics related to the article are also generated using this API.

2) Loroa - this is an audio translation API that I developed in 2021. It was a previous winner of AWS Graviton Hackathon. I modified it to synthesize an spoken audio file for each summary. This audio file is useful for visually-impaired users and can help with retention for those who learn quicker by hearing vs simply reading along. It uses Amazon Polly to create an mp3 audio file of the summarized text.

Challenges we ran into

Learning to work within the Forge framework was the most challenging aspect, as some ways of doing things are a bit different. For example, the "window" object is not available as in normal JS code, so one must learn to consult the provided documentation and develop under Atlassian's recommendations. This, of course, makes sense as these confines are in place to protect user data and privacy. Certain methods, at first frustrating, became easier once I embraced the intent and spirit of what Forge seeks to do -- make developing apps for Atlassian products a breeze while maintaining user confidence and trust.

Accomplishments that we're proud of

I believe this app will help business teams by helping team members retain and remember key information related to their daily duties. We are all working towards the common goal of increasing knowledge amongst teams and individuals. Smart Summary will hopefully be a nice add-on that lets users get up to speed much more quickly.

What we learned

This was my first time developing a Forge app. I found it similar to developing most React apps.

In researching ideas and possible concepts for my app, I got to learn more about the needs of businesses and their individual teams. Perusing the Atlassian Marketplace also helped me see what was currently available to users and allowed me to think of how I may be able to provide an offering that improves upon what is currently available.

What's next for Smart Summary

I'd like to integrate translations into this app as well. There are already Forge apps that provide translations, so I avoided including it in this version, but I'd like to explore ways on how to make Smart Summary available for different languages and regions.