This project is part of the ultimate chatbot tutorial series from TECH WITH SACH(https://www.youtube.com/techwithsach) youtube channel.
Follow the youtube tutorial to setup and configure properly.
This sample consists of a simple Dialogflow agent, a Node.js server, and a web interface that together demonstrate an approach for handling text-based conversations from a Dialogflow agent to a human operator. This also uses the Dialogflow Sentiment Analysis and makes the transition between the agent to human based on negative sentiment score.
This sample is designed as a learning tool and a platform for experimentation, not as a finished solution.
- Agent will escalate to a human operator on-demand
- Agent will escalate to a human operator after repeated failure to match an intent
- Agent will escalate to a human operator if the conversation is observed to have a negative sentiment score.
- Server routes conversation between agent and human operator depending on context
- Server supports multiple concurrent conversations between customer, agent and operator
- Customer web client supports real-time chat with agent or operator
- Operator web client supports real-time observation and participation in multiple customer conversations
- Operator web client demonstrates alerting of operator in case of customer escalation
- There is no authentication of customers or operators. In this limited demonstration, communication between users is not secure. This code is provided as an example, not a finished solution.
- Active conversations are stored in-memory with no on-disk persistence (though code could easily be modified to support this).
- Operators can only monitor and communicate with customers that have subsequently connected. Previously connected customers are ignored.
- User interface is very rudimentary.
- Dialogflow
- Node.js v6.9.1+
- Express.js
- Socket.IO
- JQuery
- Sign-up or Log-in to your Dialogflow account.
- In Dialogflow's console, select Create Agent in the left navigation and fill in the required fields and Save.
- Name for your agent, i.e.
agent-human-handoff-sample
- Select
Create
. - Go to the settings โ > Export and Import tab > Restore from zip.
- Upload the
HumanHandoffDemonstrationAgent.zip
file located in this repo.
- Upload the
- Download/clone this repo.
- Ensure at least Node.js v6.9.1 is installed.
- Within the repo directory, $ npm install to install all of the project's dependencies.
- To connect to your agent, you will need your Dialogflow Project ID.
- Back in Dialogflow console under Settings โ > General tab > copy Project ID.
- For authentication, you will need to download a service account key.
- Right below Project ID > click Service Account email address.
- In the resulting page, find the same service account email address in the table, and select the three dots under the Actions column > Create key, JSON, and then CREATE to download a key file.
- In terminal, to start the server:
$ DF_PROJECT_ID=<project ID> DF_SERVICE_ACCOUNT_PATH=<path to key file> node app.js
For example:
$ DF_PROJECT_ID=agent-human-handoff-sample
DF_SERVICE_ACCOUNT_PATH=~/Download/keyfile.json
** node app.js**
- In your terminal, you should see *Listening on :3000 and the server is now running.A.
A. If you see Error: listen EADDRINUSE :::3000
when starting the server, there is
already a process listening on port 3000 on your system. Identify and close that process, or edit app.js
to connect to a different port.
- With the server running, go to the Operator interface in your browser.
- You should see a mostly empty page with the word 'Operator' and a chat form.
- Open up a new browser tab and go to Customer interface.
- After loading, the client should immediately connect to the Dialogflow agent. You'll see the welcome message from the agent appear beneath the Customer title.
- Return to the Operator interface tab.
- You will see that a tab has appeared that represents the conversation with this customer. It will currently be empty.
- Return to the Customer interface tab.
- You can use the form to chat with the agent With Send or return.
- To demonstrate functional conversation, ask
what is your purpose?
.
- In Operator interface, you will see the history of messages between the customer and the agent. If you attempt to enter a message, you will see a warning that the customer has not yet been escalated.
- Return to the Customer interface.
- To request an operator, ask a variation of
let me talk to a person
. - The Operator interface will pop up a dialog box to alert the operator that a new escalation has occurred.
- If the Operator interface took control as a result of alert dialog, return to the Customer interface. You will see the system has responded with an introduction message from the operator. After this point, the operator can respond to the customer's messages freely.
- Still on the Customer interface, send another message (for example,
I'm having difficulty with my product
). - Open the Operator interface. You will see the customer's message. You can use the form to respond.
- Switch back to the Customer interface.
- If you responded to their message in the previous step, you will see your response in the chat log.
- In a new browser tab and go open a second Customer interface.
- In this new Customer interface, ask the agent something it doesn't understand (e.g.
I would like to book a flight
). The agent will ask for clarification. - Switch to the Operator interface.
- You will see that a second tab has appeared, representing the second customer. Click this tab to view the conversation logs.
- Switch back to the new Customer interface and enter another arbitrary statement (e.g.
I want to book a flight
).
- With the agent unable to determine the customer's intent twice in a row, it will automatically escalate to an operator.
- The server communicates with the clients via Socket.IO. It initially sends all customer messages to Dialogflow and returns the responses to the customer.
- The server monitors the
context
property of Dialogflow's response for a specific context namedoperator_request
. When it sees this context, it stops routing this customer's messages to Dialogflow and allows the operator to respond. - The Dialogflow agent can apply this context in two modes.
- In one mode, it is applied to the
Output contexts
of theOperator Request
intent, thus switching the customer to an operator when they explicitly ask for it. - In the other mode, it is applied to the
Output contexts
of the intentDefault Fallback Intent - fallback
. To see this intent in the Dialogflow interface, click the downward arrow on the intent namedDefault Fallback Intent
. Because this intent is specified as the fallback intent for the default fallback intent, it will be triggered only if the user has failed to provide a matchable input twice in a row. The first unmatched input triggers theDefault Fallback Intent
. The second unmatched input triggersDefault Fallback Intent - fallback
. This causes theoperator_request
context to be applied, triggering escalation by the server.
- In one mode, it is applied to the
You can modify this sample as a basis for your own implementation, but please be aware of the limitations expressed above.