From Bright Pattern Documentation
Jump to: navigation, search

Intelligence Engine Version 1.6 Release Notes

Enhancements

Improved Autoscoring and Autosurvey Setup

Autoscoring and Autosurvey are now managed on separate screens, accessible via the left-hand sidebar under Products. This change provides separate workspaces for evaluating agent participation in customer conversations and evaluating holistic customer experiences, making it clear which forms are assigned to which services. This separation allows you to apply distinct strategies to each.

You can manage, monitor, and set up autoscoring and autosurvey for your services directly from these pages. The interface supports bulk operations, making it possible to assign, change, or reset forms for multiple services at once. If you need to halt autoscoring or autosurvey for operational or technical reasons, you can pause and resume the process for selected services without losing your configuration.

From these dashboards, you can immediately identify the active status of your services. The dashboards also display interaction metrics over the last 7 days, including successfully processed interactions, items delayed by AI processing limits, and failures. You can download a CSV file of any configuration errors to assist with troubleshooting.

Use the Setup New Autosurvey and Setup New Autoscoring wizards to assign forms to services as desired. While you can assign a service multiple autoscoring forms as desired, a service can be assigned only one autosurvey form at a time.

Use the new Selective Past Processing wizard to summarize interactions created before ongoing summarization was configured. Past processing runs in the background, allowing you to continue working while monitoring success and error rates. This replaces the previous Recalculate and Run Past Processing functions.

Additional Language Options

You can change the display language of your Intelligence Engine instance at will. By default, language selection follows the browser’s language settings. Changing the display language automatically translates Intelligence Engine UI elements, but does not affect interaction-derived content, AI-generated text, or custom object names. Note that changing the display language may affect the date and time format.

Supported languages include Arabic and English (US).

Improved AI Intents

Designated Speaking Parties

The new Speaking Party property allows each Intent to be configured as applicable to Customer, Agent, or Any speaker. This helps improve intent detection accuracy by evaluating intents only against relevant speaker turns in real-time scenarios, and by validating returned evidence phrases against the configured speaker for post-interaction use cases such as Autoscoring, Autosurvey, and Categorization.

Defining the Number of Utterances to Resend

The new Conversation turns sent per real-time intent request setting allows organizations with Real-time Intent-based Agent Assist to define how many recent conversation turns are analyzed for real-time intent detection at a time, including the current turn.

If you use your own AI Provider Accounts for Real-time AI processing, you can adjust this setting directly.

If you use the system's default account, contact your service provider.

Improved Comparison Functionality

The Compare button on Intelligence Engine dashboards now calculates a comparison range with the same duration as the selected reporting period, immediately before it, without overlapping the selected period. For example, if 14 Days is selected from Period, the Compare button will read Compare Previous 14 Days.