Real Time Suggestions Quickstart
Real Time Suggestions provide agents with timely, context-based information directly within the Agent Desktop during live customer interactions. This helps increase agent productivity, ensure compliance, and improve overall customer experience. This tutorial explains how to create, configure, and deploy Real Time Suggestions for your agents.
At the core of Real Time Suggestions are "Intents". You configure these intents to recognize when a phrase or topic appears in an interaction that is of interest to your business. Each intent can have an associated suggestion (text, a link to an article, or both) that provides relevant information to the agent. As the conversation unfolds, the system continuously analyzes the real-time transcript. If an intent is detected, its corresponding suggestion is automatically displayed to the agent, providing immediate assistance.
Prerequisites
For users to receive real-time suggestions on any channel (voice or chat), they must be assigned the Use Real-time Intent-based Assist privilege.
Additionally, suggestions in voice services require the following:
- Agents on voice services are assigned the View real-time transcripts of voice conversations privilege.
- Call recordings are enabled for the service in either of the following ways:
- By enabling the global Quality Management setting Call Recording to record calls on all services.
- By enabling the service-specific property Record Calls on each service that will show real time suggestions.
- The following service-specific properties are active for each service that will show real time suggestions:
Procedure
Follow these steps to configure Real Time Suggestions and apply them to your services:
- Create Intents
- Refine Intent Definition
- Create and Configure a Suggestion Set
- Test the Suggestion Set
- Assign the Suggestion Set to a Service
- Verify in Agent Desktop
1. Create Intents
Intents detect specific customer or agent phrases/topics, and can be used to trigger a suggestion.
- Navigate to Intent Studio in the Intelligence Engine application, and click Create.
- The intent properties are configured from the Intent Studio:
- Enter a descriptive Name for your intent, to identify it throughout the interface.
- Intent Type: Choose between Grammar or AI intents. For Real Time Suggestions, Grammar intents are strongly recommended as they are faster and more reliable for immediate processing.
- Grammar intents are used to match keywords or phrases. They are fast to process and provide predictable matching, but require careful construction using the Grammar Intent Syntax. For example, to match phrases like "I can't login", "I cannot seem to sign in", or "I can't access my account right now", you could use the following:
CUSTOMER: "<i can't | i cannot> * <login | sign in | access my account>"
- AI intents consider each interaction in the context of a prompt written as a yes/no question. If the AI determines that the answer to the question is "yes", the intent is triggered. While potentially more flexible for nuanced situations, AI intents have higher latency as they require the conversational turns to be sent to an AI model for evaluation.
- Grammar intents are used to match keywords or phrases. They are fast to process and provide predictable matching, but require careful construction using the Grammar Intent Syntax. For example, to match phrases like "I can't login", "I cannot seem to sign in", or "I can't access my account right now", you could use the following:
- Definition:
- For Grammar intents, this is where you input the keywords, phrases, or patterns that will trigger the intent. See the Grammar Intent Syntax for more details.
- For AI intents, this is where you phrase the yes/no question that the AI will use to analyze the conversation. For example, "Is the customer expressing frustration about the recent policy change?"
- Click the + button to define the intent's matching logic (grammar or AI question) for additional languages. This allows the same intent to trigger based on conversations in different languages, each with its own specific definition.
- For Grammar intents, this is where you input the keywords, phrases, or patterns that will trigger the intent. See the Grammar Intent Syntax for more details.
- Add Precondition Intent: Select an existing intent that must be detected before this current intent is triggered. This helps create more specific and context-aware suggestions. For example, you might have a general "Refund Request" intent as a precondition, and then a more specific intent like "Refund for Damaged Item" that only triggers if the "Refund Request" precondition has already been met during the interaction.
2. Refine Intent Definition
Before deploying an intent, test whether it's triggered by the desired phrases. After saving the intent, click Test Definition to open the definition testing workspace. This example shows how to test and refine a Grammar intent, but testing an AI intent is similar.
- Click Create Test Dataset to define a sample of interactions to test your intent definition, or select an existing dataset.
- Name your dataset for future use, set the criteria for sampling interactions, and click Create.
- Once the dataset is selected, click Run Test, then select one of the interactions with "yes" in the Intent Hit column.
- The interaction transcript opens in the right pane, highlighting parts that match the intent. Use the User Answer response buttons to indicate if the intent was correctly triggered. Your answer will appear in the User Answer column.
- Use the Compare with AI option to find interactions that your grammar may have missed. Enter a question to identify interactions that conceptually match your intent, and click Ask AI.
For example, we have a grammar intentCUSTOMER: "internet" & "slow", designed to identify when the customer reports internet problems, so we ask the questionDoes the customer appear to be having problems with their internet speed? - The table of AI-tested interactions may show an example where the AI identified a match, but the intent was not triggered.
- In this case, the AI identified that the customer's statement
All the websites I visit have been taking forever to load, what is the problem?seems to indicate an internet speed issue, but our grammar intent did not trigger, since it only looks for the wordsinternetandslow. We might update our intent to something likeCUSTOMER: "<internet | site* | website* >" & "<slow | forever>"to match a broader set of statements.
3. Create and Configure a Suggestion Set
Suggestion sets group related suggestions. Individual suggestions within a set are then linked to the intents that should trigger them. Sets are assigned to specific services. A set can be used across multiple services, and any given service can have several suggestion sets applied to it.
Get started by navigating to the Real Time Suggestions page, click Create Suggestion Set, and provide a name. This name will be used to identify the set throughout the interface.
3.1 Configure a Suggestion
- In your new suggestion set, click the pencil icon
to edit the content of the first suggestion.
- Use the rich text editor to enter the agent-facing content, which can include formatted text, links, tables, and images.
- When adding or editing a link, you can configure several options:
- Open link in...: Only applicable if Auto open link is enabled. Choose to open the link in the Agent Desktop (as an iframe), a New browser tab, or a New window. Note that some sites have security settings that prevent them from being loaded in an iframe, and therefore cannot be opened in the Agent Desktop.
- Auto open link: When enabled, the URL will be automatically opened for the agent (to the target chosen by Open link in...) when the suggestion is displayed. If disabled, the agent will need to click the link to open it.
- Any number of links can be set to automatically open from a single suggestion.
- Open link in...: Only applicable if Auto open link is enabled. Choose to open the link in the Agent Desktop (as an iframe), a New browser tab, or a New window. Note that some sites have security settings that prevent them from being loaded in an iframe, and therefore cannot be opened in the Agent Desktop.
- Add more suggestions to the set by clicking Create Suggestion.
3.2 Link Intents
Each suggestion must be linked to at least one intent. Click the plus icon 
to select the intents that will trigger this suggestion.
- You can link multiple intents to a single suggestion. The suggestion will be triggered if any of the linked intents are detected.
- Click Select Intents to link the selected intents to the suggestion.
|
An intent can only be linked to one suggestion within a given set, but can be reused in multiple sets. Be advised that if you use the same intent in multiple sets that are then assigned to the same service, every suggestion linked with that intent will be displayed to the agent. | |
3.3 Add Language Variants (Optional)
You can provide translations for your suggestions to support agents and customers who speak multiple languages.
- Click the currently active language displayed at the top of the suggestion set, then select Manage Languages.
- You can add new languages or select from existing ones for your suggestion set. Once a language is added, you can switch between them to provide translated content for each suggestion.
- Provide translated content for each of the suggestions.
|
|
When a suggestion is triggered from a set configured with multiple languages, the suggestion language is chosen according to the following priority:
| |
3.4 Skill-Based Suppression (Optional)
You can configure a suggestion set to be hidden from agents who meet certain skill criteria. For example, you might have a suggestion set with basic guidance that experienced agents do not need.
- Open the kebab menu near the suggestion set title, and select Set Skill Threshold Suppression.
- Set the threshold for the selected auxiliary or service skill for which suggestions will be suppressed.
- Agents who possess the selected skill at a level equal to or greater than the specified threshold will not be shown suggestions from this particular set.
4. Test the Suggestion Set
Before deploying, test your suggestion set to ensure intents are triggered correctly and the suggestions are helpful.
- From the suggestion set editor, click Test.
- Extract a test dataset by defining the search scope, setting the number of samples to test, and clicking Run Test.
- The results table shows which intents (if any) were triggered by each sample interaction. Click on an interaction to view its transcript. Click on one of the Intents Detected to jump to the point in the transcript that triggered the intent and see the suggestion.
- You can edit the suggestion text directly from this view if needed.
5. Assign the Suggestion Set to a Service
To activate your suggestions, assign the set to one or more services.
- Navigate to the Coverage tab, Organize by Real Time Suggestions, and select Setup Suggestions from the advanced menu.
- Select the suggestion set(s) you want to assign to this service, and enable the Enable suggestions toggle.
- Click Save, and the system will start offering real time suggestions for the service.
6. Verify in Agent Desktop
Verify that the suggestions are working as expected in the Agent Desktop.
- Log in to the Agent Desktop as an agent who is assigned to one of the services configured to receive suggestions and has the permissions Use Real-time Intent-based Assist user privilege and View real-time transcripts of voice conversations user privilege.
- Initiate or handle an interaction (e.g., a voice call or chat) that is routed through the configured service.
- During the conversation, say or type phrases that you defined to trigger the intent (from Step 1).
- Observe the Agent Desktop:
- The defined Suggestion Text should appear.
- If an Auto Pop URL was configured, the article or webpage should appear on the desktop or in a new tab/window as specified.
- The defined Suggestion Text should appear.
