Sep 21, 2026
5 min read
Hear has launched a new capability that allows users to build custom reports and data visualizations using natural language prompts. This update enables CX leaders and operations managers to transform plain English descriptions into complex data queries, eliminating the need for manual spreadsheet work or technical database knowledge. By simply describing the trends or metrics they wish to see, users can instantly generate actionable insights from their customer conversation corpus.
The new Query Reports feature uses a specialized platform-mega-agent to bridge the gap between human language and technical data structures. When a user types a request—such as asking for a breakdown of customer sentiment over the last month—the system generates a validated MongoDB pipeline to pull the exact data required. This process includes a validation step to ensure the query is accurate before the report is rendered in the UI.
To further refine these insights, Hear has introduced a dedicated filter bar for these prompt-generated reports. This allows users to tweak the results without starting a new search. Additionally, the default date-range indicator has been moved into the Filters panel, where it now displays the actual date range being analyzed, ensuring total transparency in the data being viewed.
In most contact centers, generating a custom report requires a request to a data science team or hours of manual exporting and pivot-table creation. Hear removes this friction within the Analytics and Insight suite. Leaders can now ask unique questions about their operations as they arise, such as identifying specific product complaints or tracking the success of a new promotional script in real time.
This speed to insight allows managers to move directly from understanding to action. Whether it is identifying a training gap or spotting a sudden shift in customer behavior, the ability to query the conversation corpus naturally means that data is no longer a static resource, but a conversational partner.
Beyond natural language reporting, this release includes significant updates to the Compliance and QA area of the platform. A new Alert API action has been implemented, which includes a dedicated UI and API route validation. This allows teams to monitor schemas and trigger actions based on specific signal detections more reliably.
Access to the Strategy Tracker has also been expanded. Previously restricted to super-admins, this tool is now available to both Managers and Admins, allowing more leaders to oversee organizational goals. For teams handling international volume, Hear has updated its transcription routing. Arabic auto-transcription is now routed to COHERE_ARABIC, providing higher quality transcripts for Middle Eastern market operations.
Hear continues to refine the administrative experience with several key infrastructure updates. The project settings user list now features role badges—Admin, Manager, or Agent—making it easier to audit permissions at a glance. We have also introduced more flexible billing configurations, supporting contract and bracket pricing with Net 28/45 terms and customizable invoicing cadences.
On the security front, this release addresses several critical areas. We have implemented stricter organization authorization checks to prevent unauthorized disclosure of user rosters and roles. We also added a budget-exhausted banner to the Data page to provide clear communication regarding usage limits. Finally, session management has been improved with a default seven-day JWT expiration and a new setting to exempt super-admins from session revocation when necessary.
Yes, you can use the integrated filter bar to adjust the data parameters or date ranges of any report generated through a natural language prompt. This allows for quick iterations without needing to write a new request.
The Strategy Tracker is now accessible to both Managers and Admins. This change ensures that operational leaders have the tools they need to monitor compliance and performance strategies without requiring super-admin intervention.
Hear uses specialized models for different languages to ensure accuracy. For example, Arabic transcription has been updated to use the Cohere Arabic model, which provides better performance for the specific nuances of the language compared to general-purpose models.
Hear now displays a clear budget-exhausted banner on the Data page. This notification ensures that administrators are immediately aware when usage limits are met, allowing for timely billing adjustments or data management.