The Navigator Router components - Analyze voice message and Analyze digital message - are designed to detect the main topic or intent of an interaction. However, they do not capture other relevant information embedded in the contact’s message.
Navigator Attribute Detector enables customers to extract specific and multiple entities from a contact’s message, across both voice and digital interactions. These entities can include product names, locations, dates, or any other business-specific attributes.
Example: Customer says, “I’d like to book a trip on July 15th to Barcelona”. The system may correctly detect the topic as travel booking; however, the date and destination are not extracted, forcing the agent to ask follow-up questions. This increases handling time and creates friction for both contacts and agents.
Navigator Attribute Detector addresses this gap by extracting granular information upfront. This enables faster interactions, smarter automation, and richer data for analysis.
How to Configure Navigator Attribute Detector
This feature is configured directly in Talkdesk Studio through the component Detect Navigator attributes and can be used in both incoming voice and digital flows. Normally, this component is placed after the Analyze voice or digital message component, as a complement to topic detection.
For example, if a customer says, “I want to buy a set of spark plugs for my 09 Civic”, the Navigator router may classify the topic as “orders”. However, additional valuable information is present in the same message, such as the product, car manufacturer, and car model. This is where the Detect Navigator attributes component is used.
Configuration steps
- Add the Detect Navigator attributes component to your Studio flow.
- Select the input variable to be analyzed.
Typically, this is the contact’s message that was previously analyzed by the Navigator router (Analyze voice/digital message). - Optionally provide additional context.
You can specify business or industry context, such as “supplier of performance automotive and truck equipment”, to improve accuracy and ensure correct interpretation of domain-specific terminology. - Define the attributes to extract.
Each attribute requires a clear name and description.
You can extract up to five attributes per component. If more are needed, you can add additional Detect Navigator attributes components to the flow.
- Map the extracted attributes to Studio variables.
These variables can then be reused throughout the flow, such as for routing, automation, agent context, or reporting.
- Configure Studio exits. Define exits for success, no match, and error handling to control the flow behavior based on extraction results.
How agents can see extracted entities in the Conversations App
Once attributes are mapped to Studio variables in the Detect Navigator attributes component, they can be exposed to agents in the Conversations App.
To do this, navigate to “Manage Context” in Studio ({}) and select the relevant variables. These variables will then appear in the Conversations App snapshot, in the same way that Navigator router information is displayed today.
When an agent receives a call or digital interaction that passes through this component, they can immediately see key details extracted from the contact’s message. This eliminates the need for contacts to repeat information and allows agents to act faster and more confidently.
The extracted data can also be used in automations and reporting. This ensures that teams have the information they need upfront, enabling better decision-making, improved analytics, and more proactive business actions.
To learn more about eligibility, pricing, and activation of this feature, please contact your Talkdesk account representative.