Aug 31, 2026
4 min read
Hear now allows you to link external AI assistants like Claude directly to your workspace to query customer conversation data without manual exports. This integration enables sales, marketing, and CX leaders to access deep insights from calls and chats within the AI tools they already use, accelerating decision-making and streamlining workflows.
Hear uses the Model Context Protocol (MCP) to create a secure bridge between your customer conversation corpus and external AI environments. By enabling this feature, managers can use an assistant like Claude to retrieve specific data points from the Analytics and Insight area of the Hear platform. This allows for natural-language querying of complex datasets, such as identifying recurring customer pain points or summarizing sentiment across hundreds of recent calls.
The integration is designed to respect the existing security architecture of Hear. When a user queries data through an external assistant, the platform enforces role-based data scoping. This ensures that sensitive information remains protected and that users only access the logs, calls, and reports they are authorized to view within the native Hear interface.
CX and operations leaders often rely on a variety of specialized tools to manage strategy and performance. Previously, moving data from a contact center intelligence platform into an external LLM required manual effort, which created friction and delayed insights. With this release, the intelligence gathered by Hear becomes a portable resource. Whether you are drafting a strategy memo in an AI document editor or analyzing market trends in a chat interface, your Hear data is now available to provide evidence and context.
Beyond the AI assistant integration, this release includes several updates to the Quality and Coaching and Compliance and QA areas of the platform to improve accuracy and user experience.
We have introduced an Audio Boost feature to improve the clarity of customer recordings. To further assist with coaching, transcripts now synchronize with audio playback, highlighting text in real-time as the recording plays. This makes it easier for managers to review specific moments in a conversation and provide targeted feedback.
The Hear reflection agent has been updated to support scheduled AI tasks, allowing for more consistent monitoring of customer signals. We have also improved the feedback loop for conditional alerts; when a manager provides feedback on a signal, the system can now suggest direct edits to the signal logic rather than defaulting to a prompt-based change. This ensures that your automated QA processes remain precise and easy to maintain.
To support larger organizations, we have hardened our team management actions and introduced proactive inactivity warnings for user sessions. Administrators can now deactivate user access more efficiently, and the platform now supports automated monthly invoice generation for opted-in organizations via Stripe and Green Invoice.
You can connect any AI assistant that supports the Model Context Protocol, such as Claude. This connection allows the assistant to securely read and analyze data from your Hear workspace based on your specific queries.
The integration is enabled for all Hear accounts. However, for security reasons, it is restricted to users who hold a Manager permission level or higher.
Yes, Hear enforces role-based data scoping for all external queries. The external assistant can only access the specific calls, logs, and reports that your Hear user profile is permitted to see.