Stronger Data Governance, Deeper NiCE CXone Integration, and More with Cognigy.AI 2025.25 Release

Cognigy.AI 2025.25 introduces targeted enhancements focused on privacy-aware data governance, native integration  with the NiCE CXone Agent Assist Hub, and more efficient collaboration between frontline agents and supervisors.

Native Data Redaction for Privacy-Aware Logs and Analytics 

Data protection is a fundamental requirement for enterprise AI. With the new Data Redaction feature, Cognigy.AI can automatically detect and redact sensitive information, including personally identifiable information (PII), from both conversation logs and analytics data.

Why It Matters

This capability represents a significant upgrade over existing Blind Mode and Endpoint-level data masking options, which obscure entire logs and effectively disable analytics visibility. Instead of hiding all data, Data Redaction selectively removes only sensitive elements, preserving the surrounding conversational context and analytical value.

For teams operating in regulated environments or under strict privacy requirements (such as GDPR or CCPA), this feature strengthens data governance while reducing operational trade-offs. Logs remain readable, analytics remain actionable, and sensitive information is never exposed, enabling compliance without sacrificing insight.

See the Data Redaction Demo in Q4 Cognigy.AI Tech Update

Key Advantages

• Automatic, pattern-based detection and redaction of credit card numbers, phone numbers, IP addresses, social security numbers, and more
• Custom redaction definitions and data types for domain-specific or organization-specific sensitive data, offering full control and flexibility

Enabling Cognigy Agent Copilot in NiCE CXOne Agent Workspace

The 2025.25 release also extends Cognigy.AI’s integration footprint with the NiCE CXone Agent Assist Hub (AAH) Endpoint, now fully available to all users.

Why It Matters

This endpoint enables human agents working with the NiCE CXone contact center platform to access Cognigy Agent Copilot directly within their native UI. It also supports a dedicated Agent Copilot deployment model, where end customers interact exclusively with human agents, while AI operates entirely in the background.

In this setup, Cognigy Agent Copilot assists agents with real-time guidance, recommendations, and contextual insights, without exposing AI Agents to the customer-facing experience. This allows organizations to enhance agent performance and consistency while maintaining full control over how and where AI is applied. 

See the NiCE CXone AAH Endpoint Demo in Q4 Cognigy.AI Tech Update

Escalation to Supervisors for Live Agent Conversations

Cognigy Live Agent now introduces the Escalate to Supervisor feature, enabling human agents to escalate active conversations directly from the Live Agent interface to a designated supervisor.

Why It Matters

This feature is particularly valuable in scenarios such as customer complaints requiring managerial approval, billing disputes, policy exceptions, or emotionally charged interactions where de-escalation expertise is required. Instead of relying on manual coordination or side channels, agents can escalate conversations seamlessly and in context.

Supervisors can take over the conversation with full visibility into the interaction history, improving resolution times while providing better real-time support for frontline agents. The result is a more structured and reliable escalation flow that enhances service quality, agent confidence, and overall operational efficiency.

See the Escalate to Supervisor Demo in Q4 Cognigy.AI Tech Update

Other Improvements

Cognigy.AI 

• Introduced the Send Email and Handover to AI Agent tools, enabling AI Agents to send emails directly and transfer conversations to other AI Agents within workflows
• Added the Tenant ID field and moved the Billing Time Zone to the Change Organization Information section in the Management UI
• Removed the gpt-3.5-turbo, gpt-3.5-turbo-instruct, and gpt-4 models
• Replaced the request-promise library with Axios for HTTP requests sent from the transformers
• Added support for custom hostnames when deploying the platform behind a CDN or firewall
• Updated logic for generating session IDs for Endpoints. Now, each Endpoint and third‑party account combination, for example, a WhatsApp Endpoint and a WhatsApp account, gets a unique session ID with no duplicates across Endpoints and Projects
  

Cognigy Voice Gateway

• Improved connection handling when multiple WebSocket connection attempts are made, reducing dropped and failed connections
• Improved handling of DTMF signals and Barge-in scenarios, making voice interactions more responsive and accurate

Cognigy Live Agent

• Introduced Live Agent OData Endpoint version 2.0, featuring enhanced functionality and a new teams collection for managing team data.
  

Cognigy Insights

• Added Markdown rendering support to the Transcript Explorer
• Updated the session’s Locale information when switching Locales in the Flow