Tracking code installation
Tracking page visits
Intents and Entities
Building Dialogflow agent and connecting to Activechat
Sending user input to NLP engine
Triggering bot skills by intents and events
Working with entities
NLP best practices
Sending order updates
Real power of Activechat Visual Bot Architect comes with external integrations. In a couple of mouse clicks you can connect powerful natural language understanding engine, e-commerce shop, CRM etc.
Dialogflow is a free AI service from Google that provides powerful natural language understanding capabilities. We have direct integration with this service through API, so you can teach your chatbot understand natural language in a couple of simple steps. Working with NLP engine is quite complex topic, and we also have a whole educational course about it (check “Natural Language Chatbots” in this Academy). So we would not dive into details immediately, and will just show you how easy is the connection process.
First of all, you need developer access key to your Dialogflow agent. Go to Dialogflow console and click agent settings. Find “Developer access key” below and click to copy it.
Now go to “Integrations” tab in your bot settings in Activechat and select “Dialogflow” on the left. Finally paste your developer access key and click “Connect”.
As a result every intent that you have in Dialogflow agent is assigned an event in Activechat, and you can use CATCH blocks to start specific skills based on these intents. To send user input to Dialogflow you should use special NLP block. Here is the most common setup for this in “default” skill: