Inspiration
I could not find the skill that reads from user-defined list of rss and podcasts, and also I wanted the skill that can summarize and translate the rss items.
What it does
Right now rss-reader just reads rss items and podcast items crawled from the rss feed urls you have defined. And it read in a chronological order.
How we built it
All of the authentication and front-end is served by express application on elastic beanstalk. OAuth server is also served by express application, but the actual Alexa request is handled by lambda. Right now I'm using mongodb database which is served on mlab. But I will change the database to DynamoDB which can be accessed directly from Lambda functions which will make two layers ( Lambda, Express ) to share the same database.
Challenges we ran into
The biggest challenge was to find the right api to extract article. Most of the article extraction APIs are very pricey and does not provide the right features I wanted such as summarizing article contents. Right now I'm using Embedly, but I'm looking at Aylien to replace Embedly.
Accomplishments that we're proud of
The best accomplishment is that our service is very stable and it surely does its job to help people save time to actually go to the news site to read articles.
What we learned
- I learned how to create a modular platform that is less prone to errors and crashes.
- I learned how to analyze people's interest by analyzing the content each user is interested into sentiments.
What's next for RSS Reader
Following are the list of features I want to add to RSS Reader.
- Use collected sentiments data to suggest trending articles to the user.
- Summarize the article.
- Translate the article after detecting the language of the article.
Log in or sign up for Devpost to join the conversation.