Dialogflow recently recognized XAPPmedia as one of only five agencies listed on its website. We are proud of joining this exclusive group in large part because we really like using the product. One great example where team XAPP put Dialogflow to good use is the award-winning Bravo Tango brain training Google Assistant Action we built in collaboration with 360i for National Geographic.
Many developers are building all of their voice apps fully inside their code base. Dialogflow lets you separate the voice app logic from the content and intents. This makes the build process more efficient, less reliant on technical resources and offers additional tools to improve the user experience.
Separating Logic from Content and Intents
For Bravo Tango, we used Dialogflow to define the intents and content that drove the user experience. That work was completed by non-technical team members and allowed our engineers to focus on building out the logic on XAPP’s One Voice AITM platform for voice apps. Then we were able to use Dialogflow post-launch to update the Natural Language Understanding (NLU) and improve the user experience. My colleague Andrew Gassen commented:
“All of our non-logic elements for Bravo Tango, such as intents, utterances and NLU tuning, were done in Dialogflow. It was used to create the content and user experience while the engineering team focused on the logic. Without Dialogflow we wouldn’t have been able to move as fast because we would have bottlenecked at engineering. After launch, we were able to use Dialogflow to see what specifically people were saying when using the Google Action and how it was mapped to a response. We could then audit whether those responses were optimal and update the model based on real user input as opposed to our guesses.”
Using One Voice AI and Dialogflow Across Multiple Assistants
Bravo Tango was only deployed on Google Assistant. However, we are using Dialogflow in voice apps published across multiple voice assistants as well. A key benefit of the One Voice AI platform is simultaneous multi-assistant publishing of voice apps. We have tools for content and logic management across the platforms and then are able to use Dialogflow for the cross platform NLU tuning. So, it is a really good fit for the scope and scale of the projects we are launching for leading brands ranging from Progressive Insurance to E. & J. Gallo.
We definitely appreciate the acknowledgement by Dialogflow of our expertise in launching and maintaining voice apps. As the only company with more than 1,000 voice apps launched and managed on Amazon Alexa and Google Assistant for brands and media companies, we have learned a lot and refined the tools and processes we employ. Dialogflow was one of the few that made the cut and we really value the added benefit it brings to our voice app development and managed services. We also liked that the recognition offered a good excuse to provide additional insight into how voice apps are built and maintained.
Google Assistant Webinar for Brands
You can learn even more about using Dialogflow and building for Google Assistant in our upcoming webinar next week.
Topic: What Goes Into Building an Award Winning Google Assistant Action, Bravo Tango Case Study and Best Practices from Leading Brands
Date: June 19, 2018
Time: 1:00 pm EST / 10:00 am PST for 30 minutes