Using Jargon to Adopt a Skill for Amazon Echo Auto
Amazon has started fulfilling orders for the Echo Auto device. This will open up a whole new category of voice first functionality for customers. I recently used the Jargon SDK to help adopt my Restaurant Finder for this device. This skill allows customers to find and filter restaurants based on Yelp reviews. You can ask for types of cuisine such as Chinese, or qualifiers like great, cheap, or open. You can specify an American or Canadian city to find restaurants, or use the device’s registered ZIP code.
Before writing code, I wanted to update and simplify my content for automotive use. The current interaction flow sounds something like this:
Alexa, ask restaurant finder to find cheap Chinese food in Seattle
I found 50 cheap Chinese restaurants in Seattle. Say read list to start reading the list or filter your search with additional conditions like great.
Reading 5 restaurants. You can ask for more details by saying the corresponding restaurant number or say “more” to hear more results. 1 Zheng Cafe. 2 Mike’s Noodle House. 3 Xi’an Noodles. 4 King Noodle. 5 Qq by Facing East.
Zheng Cafe is located at 230 6th Ave N in Seattle. The phone number is (206) 588–2882. It is currently closed. See the Alexa app for more details. What else can I help you with?
This ends up being too wordy for drivers, requiring concentration to hear the full list of restaurants and details. Rather than forcing a customer to go through a list I wanted to start by reading the closest option. I also wanted to keep the flow simple, relying on natural responses to navigate the results. I decided to change the interaction to sound like this:
Alexa, ask restaurant finder to find cheap Chinese food near me
Mike’s Noodle House is 1.2 miles away. Would you like to hear about it?
Zheng Cafe is 1.5 miles away. Would you like to hear about it?
Zheng Cafe is located at 230 6th Ave N in Seattle. The phone number is (206) 588–2882. What else can I help you with?
I wanted to provide these different experiences with minimal changes to the code. Fortunately, these responses pass through the same state in my code. All I had check in my code is the device type to use a different formatting string, and Jargon took care of the rest. I was able to set the same set of variables in my call to Jargon (Phone Number, Name, Distance, Location, and Phone) even when some of them aren’t used in a given context. With Jargon, that’s OK. And now my content is more readily maintainable if I want to make additional changes to update the response for auto or non-auto devices.
Read about the other steps I went through to enable Restaurant Finder for Echo Auto, including code samples for Jargon integration and other changes, in my Medium post.
About Garrett Vargas: Garrett Vargas is CTO at CarRentals.com, part of the Expedia Group. He has over 20 years of engineering leadership experience in technology. Over the past few years, he has published more than a dozen Alexa skills ranging from casual games to a skill for CarRentals.com customers to manage their bookings. For more voice related posts by Garrett Vargas, visit his blog at medium.com/@garrettvargas
Looking to get started with the Jargon SDK? Check out these additional resources below:
Get Started with Jargon
Jargon SDK Starter Templates
Comments or questions? Contact us at firstname.lastname@example.org