A little about Action On Google
Hey, flocks!
So, you’re here to know about Actions On Google! First, let me go through some simple terminologies. I hope you all are very familiar with the Google Assistant. Personally, I love her so much!
You know, the Google Assistant was launched in the io16, the annual developer conference conducted by Google. Google Assistant was introduced to help users across the world right when they needed. Just talk with her by saying, “Hey Google”! Exact one year after, on the same stage, Google announced a platform for third-party developers to build Actions for the Google Assistant, to get things done for the users.
So, what exactly the term “Action” means? It’s a kind of app, a conversational app you build for the Google Assistant. For example, think about you’re asking the Google Assistant about the weather, “Hey Google, what’s the weather right now?” then the Google Assistant will trigger the Weather Action built by Google and respond you with the Weather data.
Actions are super amazing, you can handle your customers with your Google Assistant. Think that you’re running a bike repair shop, instead of handling a lot of phone calls or messages to handle the appointments for your clients, what if you have a Google Assistant Action. The total game will be changed. Your clients can talk with the Action and make appointments at any time as you set. They just have to start talking by something like, “Hey Google, talk to Peter’s Bike Repair Shop” (here ‘Peter’s Bike Repair Shop’ is the Action name).
Google gives you two handy platforms to build Actions. One is the Actions On Google Console, where you can manage your Actions’ Assistant Directory setup, view your Actions’ analytic data and more. The second is the Dialogflow console. It lets you build conversational or voice apps, chatbots etc. You can integrate your app with Google Assistant or export it as a chatbot on Telegram, Twitter, Facebook Messenger etc.
Let’s now take a look into how it works!
First, the user says something as the user input to the Google Assistant in natural language. The Google Assistant will parse it into some conversational API request, say some complicated JSON model data objects and finds the best-matched Action and feed the data to it. Now it’s your Action’s turn. Your Action will create some response as you build using the tools like Dialogflow and send it back to the Assistant. Then the Assistant will resolve those conversational API responses to Action response in natural language, and respond back to the user.
Next, Dialogflow!
Dialogflow gives us some super amazing features including ‘Intent Matching’, ‘Entity Extraction’, and ‘Fulfilment’ to build high-quality conversational apps. Intent matching is simply the process of matching and categorizing user utterances to an intent. Dialogflow uses Machine Learning to match user utterances. Let’s see an example, assume that your action will accept a color name and give an appropriate response. Users can say the same thing in many ways like, “my favorite color is red”, “I love black”, or simply “blue” so on and so on. All these utterances will trigger just one intent, let’s name it the ‘fav-color’ intent. And the entity extraction is the process of identifying the keywords. For example, I just said some example utterances, all these contain a color name. Here the color name is an entity. And finally fulfillment. The most powerful feature of Dialogflow. Using fulfillment you can connect your own back-end script to respond to the user appropriately. Fulfillment uses the Cloud Functions for Firebase.
So, how will you create your own fulfillment? Google provides a node.js client library for you to build your own fulfillment. You just have to import modules and use.
And, that’s it!
One more thing, we’ve got a super amazing Developer Community Program to support your efforts in building great Actions for the Google Assistant and to celebrate your accomplishments. The best part? If you’ve submitted an eligible Action, you’re already a member. You can learn more about the Developer Community Program from this video:
There we go! I’m getting excited to see what you build. You can also, follow us and join the conversation using the hash-tag #AoGDevs on Twitter or any other social media!
Thanks!
