The amazon echo takes short pieces of text as input. It seemed a natural fit to make it work with twitter's short format in order to help for applications in which keeping hands free is important, or for the blind.

What it does

It translates any audio given to it into a tweet which it then sends.

How I built it

We wrote java aws lambda instance.

Challenges I ran into

The amazon echo is not particularly good at recognizing free strings, instead it prefers that there be an expected result. Therefore, we had to build a database of tweets in order to hone in its accuracy.

Accomplishments that I'm proud of

Making the echo work with an external API was challenging but we are very proud to make these two complicated systems work together.

What I learned

We learned that the Echo will not successfully transcribe strings without a database of some kind to help its filters. We also learned more AWS skills and the twitter API.

What's next for TweetingEcho

Support for more commands such as reading your timeline and direct messaging.

Share this project: