We were inspired by the MachAlliance theme - which is getting unstuck. What better way to "unstick" people than letting them interact with our API in a brand new way? We were particularly interested in the kind of impact this could make for those that might not be able to access a computer as easily because of a visual impairment. Our original plan included adding a multi language component to this but we ran out of time.

What it does

The project is a fully hosted solution that other implementors can use to integrate OrderCloud with Alexa simply by configuring some things in Amazon (no code required) however the full source code is also available for anybody that wants to take it and run with it and build their own custom queries. It has all of the boilerplate you'd need to make a really great solution.

How we built it

Though most of the code is in the backend there is also a login screen used for authenticating into OrderCloud. A lot of the stuff was built by using amazon SDK's such as the amazon ask sdk.

Challenges we ran into

Getting authorization to work ended up taking more of our time than we had initially expected. Just understanding how to get things done in Alexa was challenging because no one on our team had worked with that technologies.

Accomplishments that we're proud of

Finally getting OrderCloud talking to Alexa! Getting a response from Alexa based on API calls from OrderCloud was the main goal and also the main accomplishment.

What we learned

With no one having worked with Amazon Alexa before, we learned how to integrate with voice assistants, all the authorizations to hook Amazon Alexa up to Ordercloud, and how to host those authorization workflows through Heroku. Along with that, we also learned more about voice UX, ContentStack, and other MACH technologies.

What's next for OrderCloud Voice

With the Ordercloud integration done, that opens up a lot more possible skills for Amazon Alexa to learn. These are a few of next steps we had planned:

  1. Multilingual support
  2. Branch out to multiple voice assistants (Google Voice, Siri)
  3. Earnings between two specific days (instead of "yesterday" or "this month", we could do "between January 1st and January 15th")
  4. Leverage ContentStack for storing login messaging and logos per client

Built With

Share this project: