Twitter Effect

Inspiration

We wanted to be able to interact with Twitter using just our voice in a natural way. Since Google Assistant couldn't do it we decided to do it ourselves.

What it does

• Reads tweets from specific people from anyone based on their displayed name or twitter handle.
• Reads tweets based on a specific category such as sports, music, or funny tweets.
• Reads tweets from your own timeline.

How we built it

We used Watson to analyze tweets given from the Twitter API, sorted, filtered and parsed throughout to find specific keywords and matched key user inputs and bridged it all together with Google's api.ai/Dialogflow to use with Google Assistant.

Challenges we ran into

Managing asynchronous code in Node.js was one. Integrating multiple API's effectively and efficiently in our allocated time was another.

Accomplishments that we're proud of

• Starting out with just the idea of what wanted to make
• Focusing on the most important aspects first.
• Maximizing the time available by working for the majority of the available time.

What we learned

Spending the extra time to set up things and plan more will go a long way.

What's next for Twitter Effect

A complete hub for a google skill and twitter collaboration. Send your own tweets by voice. More dynamic features such as getting tweets from intuitive custom categories such as family or co-workers. Reading mentions. Getting the most liked or retweeted tweet of the day.

Built With

• dialogflow
• express.js
• google-console
• ibm-watson
• javascript
• node.js