Setup

Can I turn off all recommendations?

Yes, you can turn off recommendations globally. To do this:

  1. Access Conversation Assist.
  2. Click Settings from the menu at the top.
  3. Under General, click the Suspend toggle.

Can I turn off recommendations from an individual recommendation source?

Yes, you can do this per recommendation source. To do this:

  1. Access Conversation Assist.
  2. Click Recommendations Sources from the menu at the top.
  3. Click the Knowledge Bases tab or the Bots tab, as appropriate.
  4. Beside the source, click the Pencil icon (Manage source) icon.
  5. In the Manage recommendation source dialog, change the Status to “OFF.”
  6. Click Save.

Why aren’t my knowledge bases discovered?

There are a few reasons why this might be the case. See our troubleshooting info.

Agent experience

Why aren’t my knowledge base answers formatted?

Currently, only hyperlinks are supported.

When I join a bot a conversation, the bot doesn’t always begin the flow that I expect. Why is this?

When you delegate a bot to a conversation, the specific bot flow that is triggered depends on the consumer’s most recent message. This is the case for bots added to conversations using inline recommendations and also for bots added using the On-Demand Recommendations widget.

How are recommendations made?

To make recommendations, Conversation Assist analyzes the available bots and knowledge base articles that match the consumer’s intent, and it finds the best ones. The recommendations are made by choosing those ranked highest by relevance score.

Under the hood, Conversation Assist automatically searches a knowledge base using the KnowledgeAI search offering.

The rules for how the recommendations are made and ordered are as follows:

  1. Which recommendations to include? First, include all answer (article) recommendations. Each eligible knowledge base is queried individually for the single, highest-scoring article match.

    Second, include all bot recommendations. This means that answers are included before bots even when the answer scores are lower than that of the top bot recommendation.

    The scores for included answers must meet the Answer Confidence threshold, which is configurable during setup. Similarly, the scores for included bots must meet the Bot Confidence threshold, which is also configurable during setup.

  2. How to order the recommendations? Within each subgroup of recommendations (answers, bots), sort the recommendations by relevance score in descending order so that the higher the score, the higher the recommendation.

    As an example, assume that you have set up 2 knowledge bases and 3 bots. They get called for recommendations, and the following scores are returned:

    • Article in knowledge base 1 = 90%
    • Article in knowledge base 2 = 60%
    • Bot 1 = 100%
    • Bot 2 = 80%
    • Bot 3 = 0%

    If the account’s Conversation Assist settings are 1) maximum number of recommendations = 4, 2) Answer Confidence = GOOD (70%), and 3) Bot Confidence = 70%, then the agent receives the following, ordered list of recommendations:

    • Article in knowledge base 1 (90%)
    • Bot 1 (100%)
    • Bot 2 (80%)

    Knowledge base 2 and Bot 3 have been filtered out because their relevance scores didn't meet the confidence thresholds that were configured.

    As a second example, assume that you have set up 4 knowledge bases and 2 bots. They get called for recommendations, and the following scores are returned:

    • Article in knowledge base 1 = 72%
    • Article in knowledge base 2 = 80%
    • Article in knowledge base 3 = 100%
    • Article in knowledge base 4 = 75%
    • Bot 1 = 100%
    • Bot 2 = 100%

    If the account’s Conversation Assist settings are 1) maximum number of recommendations = 4, and 2) Answer Confidence = GOOD (70%), and 3) Bot Confidence = 70%, then the agent receives the following, ordered list of recommendations:

    • Article in knowledge base 3 = 100%
    • Article in knowledge base 2 = 80%
    • Article in knowledge base 4 = 75%
    • Article in knowledge base 1 = 72%

    In this case, the agent doesn’t see any bot recommendations because article recommendations are always included first. The maximum number of recommendations is 4, and there are 4 article recommendations with a relevance score of GOOD or better.

Agent feedback (thumbs up, thumbs down)

Is the feedback of my agents captured in the reports that are available for download?

No, not at this time.

No, not at this time.

No, not at this time.

Is agent feedback used to enhance the underlying recommendation engine?

No, not at this time.

Metrics and reporting

On the Home page, in the Recommendations widget, the Summary data does not align with the graph data. Why is this?

There’s a bug in the Summary information in this widget. It should show the number of used recommendations “over” the number of offered recommendations, where the latter (the denominator) reflects the recommendations that were and weren’t used.

Currently, the denominator is incorrect. It reflects only the recommendations that weren’t used.