First of all, I’d like to express how thankful I am for discovering Rasa, which is a great platform to build chatbots.
I’d love to use it in my organisation, whose communication solution is Google Workspace. I’m new to chatbot development and I could not gather from the docs at Google Hangouts Chat how to indicate in the Hangout’s API side where to point to my rasa server. Furthermore, I guess there are several different alternatives from both sides to establish pass messages back and forth between Hangouts API and Rasa server, so I’d love some directions where to learn in more detail how to better leverage Hangouts API and Rasa server.
Does anyone know of any tutorial or lost documentation instructing on how to set up the communication between a Google Hangouts chatbot powered by Rasa server?
Hi everyone,
I also definitely need help. I am trying to follow the doc Google Hangouts Chat but it is not helpful. And unfortunately, there is no much tutorial about deployment of Rasa to Google hangout. Any help would be appreciated.
I posted this thread a couple of weeks ago and unfortunately I’ve got no answer. I’ve digged deep and finally I found a Telegram group of Rasa enthusiasts here in Brazil where people could help me out.
Fortunately, the solution was pretty straightforward: after having created a project in Google Cloud Platform, first you must enable the Hangouts API. I guess you’ve reached this point.
Make sure that your bot is live and in the connection settings part I selected “Bot URL”. In the “Bot URL” field that will appear, you can inform your hangouts webhook url. After enabling hangouts in the endpoints.yml file and running rasa, it should be something like:
https://.ngrok.io/webhooks/hangouts/webhook
Save your configuration and voilà! At least it worked for me this way!
Hi @otmarjr !
Thank you so much! your explanations were extremely helpful. I followed all steps and found my bot in google chat. But when I sent a message it says “not responding”. Did your chatbot start responding right after you saved the configuration? Also, did you make any changes in the credentials.yml? I deployed my chatbot to FB before and it was requiring changes in the credentials.
Yeah, you need to change your credentials.yml file to enable the hangouts channel, you must add an uncommented line to your credentials file:
hangouts:
I’m also in my bot’s development phase, so I’ve not yet configured any password or token for Hangouts.
With regards to the “not responding message”, I remember it happening for the first few tries with my chatbot. I was hosting from my local machine using ngrok, and then I noticed an error message in the console output produced by ngrok while it’s relaying my local rasa server. I just fixed whatever it was complaining about, but I don’t recall exactly. I guess now it’s time for you to look for error messages in your Rasa server log files or whatever you’re using to host it.
No, I’m not using Google cloud platform to store my data. I’m still in the early stages of my bot, but I soon I hope to make it available to some of my peers to evaluate using Rasa X. I intend to install Rasa X in our network and use it store conversations for further analysis in our infrastructure.
Hi again @nik202
When I deployed my chatbot to Rasa X or Facebook Messenger, these channels stored the conversations between the bot and the user. However, when I deployed the bot to Google Chat, I couldn’t find any stored conversations in Google Cloud. Is google chat able to store this data somewhere?
Thanks!
Hello @seymanur and all other interested in Rasa’s Google Chat integration. I’ve commented in a blog post about Rasa and Petr Janik kindly created a very instructive and detailed tutorial! Check it out at RASA - Google Chat integration - DEV Community
