Monitoring Bots

Master the art of creating and managing monitoring bots with Khoros Care

Live bots are a great way not only to lighten the load on your live agents but to guide the user to the answers they need quickly and reliably. They perform important tasks, including responding to user questions and performing repetitive tasks such as account creation, information lookup, and more. They are versatile and invaluable resources. However, they are limited to a single deployment per source integration.

So, what can you do when you need more than one bot working with a given source integration? Monitoring bots are Khoros' powerful answer to this very question.

Monitoring bots enable you to create passively-monitoring bots that can perform administrative tasks, utilize context sourced from external resources (content management systems, databases, etc.), and do so in parallel with a live chatbot, agent, and/or additional monitoring bots.

Monitoring bots empower you to give stakeholders within your organization the ability to incorporate their respective resources into a Khoros Care conversation without having to plug them directly into the live chatbot.

This modular capability gives companies a seamless method of maintaining information security without sacrificing the customer or agent experience.

📘

Note

As of September 2022: Monitoring bots can listen in for contextual information on both sides of the conversation, including conversations involving agents, live bots, and the end user.

What can monitoring bots do?

Like live bots, monitoring bots listen to events throughout the duration of the conversation. They receive updates from the moment the conversation is initiated to the moment it is over. Unlike live bots, monitoring bots have a more restrictive set of functionality, preventing the chance of conflicting messaging between the user and live bots or agents.

This chart highlights some of the differences between live and monitoring bots:

Bot Actions/Mode Live Monitoring
Reply with message Yes No
Typing start Yes No
Typing stop Yes No
Authenticate Yes No
Conversation control Yes No
Author attributes Yes Yes
Note Yes Yes
Priority Yes Yes
Tag Yes Yes
Resolve Yes Yes
Workqueue Yes Yes

The functionality of monitoring bots is limited to more supportive tasks such as context enrichment and supporting the conversation flow. They can listen for specific keywords, compare contextual information against external resources (CMS, database, etc.), and apply background changes to the conversation.

Here are the things a monitoring bot can do, along with example use cases:

  • Author Attributes: Identifying a user as a Platinum Member and applying the appropriate attribute(s) to their user profile in Khoros Care.
  • Notes: Document actions it took (and reasons for them) to provide additional context for agents.
  • Priority: Adjust the priority of the conversation. This is useful in cases where elevating conversations that meet specific criteria is necessary.
  • Tag: Tag conversations based on contextual information provided by an external resource (CMS) that has more information about the user than is necessary to provide to the agent/live bot.
  • Resolve: Resolve conversations once specific criteria have been met. For example, when a ticket is created or there is a successful over-the-phone resolution with a call center representative.
  • Workqueue: Directly assign a conversation to a work queue. For example, if your company has an event going on, and your CMS identifies the user as an attendee, you can direct the conversation to a specific work queue set up for attendees.

Registering a New Monitoring Bot

Registering a new monitoring bot is done very much the same way as a maintenance or live bot. The only difference comes in the registration payload.

The typical registration payload includes a mode parameter. Instead of setting the value to MAINTENANCE or LIVE, you would set the mode to MONITORING.

Here's an example registration payload:

{
    "platform": "LITHIUM",
    "companyKey": "[COMPANY KEY]",
    "networkKey": "apple",
    "externalId": "example-e66a-11e7-86ec-b1a405106b73",
    "appId": "my-abc-bot",
    "name": "My Apple Business Chat Bot",
    "callbackUrl": "[CALLBACK URL]",
    "mode": MONITORING,
    "credentials": {
        "type": "HMAC | BASIC_AUTH",
        "identity": "[USERNAME OR HMAC KEY]",
        "secret": "[SECRET]"
    },
    "contact": {
        "email": "[EMAIL ADDRESS]"
    },
    "avatarUrl": "https://www.example.com/bot_avatar.jpg"
}

Limitations

With monitoring bots, you are no longer limited to a single bot per source integration. You can deploy a total of 10 bots for each source integration. The total live bot maximum remains at 1, but it can be combined with up to 9 monitoring bots. If you aren't using a live bot, you can deploy 10 monitoring bots.


Did this page help you?