Sera AI

Sera AI is the intelligence layer embedded across the Xurrent platform. Rather than a standalone chatbot or bolt-on module, Sera AI is an AI fabric: a mesh of capabilities woven into the platform that work together to reduce manual effort, accelerate resolution, and improve accuracy. It powers request classification, knowledge article creation, request summaries, sentiment analysis, automation building, Note Assist, and the Virtual Agent.

Enabling Sera AI

Sera AI features are controlled by the Xurrent +AI toggle in the Account Settings section of the Settings console. An account owner or account administrator can enable or disable AI features for the account.

Once enabled, AI-powered features become available across the platform. These include:

• Request classification (auto-suggest or auto-apply)
• Request summaries
• Knowledge article creation and improvement suggestions
• Sentiment analysis
• AI Automation Builder
• Note Assist
• Virtual Agent (requires additional setup in Self Service Settings)

To control which parts of the organization have access, AI can be enabled or disabled at the support domain level. This allows organizations to roll out AI features incrementally or restrict them to specific support domains.

Suggest vs. apply

For request classification, the Account Settings provide two modes:

• Suggest: Sera AI proposes values for Category, Impact, Service Instance, and Team, but a specialist must confirm before they are applied.
• Apply: Sera AI automatically sets these field values without requiring manual confirmation.

Organizations that want to validate AI accuracy before fully automating can start with Suggest mode and move to Apply once confidence is established.

Language support

All languages available in Xurrent personal preferences are supported. Xurrent supports over 60 languages. Sera AI keys off each individual user's personal preference language setting to determine how content is surfaced back to them. This extends to the Virtual Agent experience, Note Assist suggestions, and all other AI-powered features.

Content that was not originally written in the user's preferred language is auto-translated as part of the standard Xurrent translation capability.

Data privacy and security

Sera AI is built on Amazon Web Services (AWS) Bedrock. Xurrent operates a RAG (Retrieval-Augmented Generation) model, combining generalized information from the underlying foundation models available in AWS Bedrock with permissioned data within the customer environment.

Key security guarantees:

• No training on customer data: AI training data is never stored. Customer data is never commingled between organizations.
• In-region processing: All processing is maintained in customer regions within AWS Bedrock. For organizations operating under GDPR, ITAR, or similar regulatory frameworks, AI processing remains in-region.
• No data retention: In-session calls are never stored. There is no lasting transactional record within the broader ecosystem.
• Permissions respected: All existing Xurrent permissions are respected. AI features do not expose data that a user would not otherwise have access to.

Sera AI has been deployed with many customers across European markets and has undergone rigorous compliance reviews. The Xurrent Trust Center provides publicly available resources about responsible AI practices.

The AI fabric concept

Sera AI is best understood as an interconnected set of capabilities rather than individual features working in isolation. For example, when the Virtual Agent creates a request, auto-classification immediately routes it to the right team, sentiment analysis evaluates the user's tone, and a summary begins building as notes are added. These capabilities are woven together to reduce handoffs and accelerate triage.

Getting the most from Sera AI

To get the most business value from Sera AI:

• Knowledge base: A well-maintained knowledge base is central to virtual agent effectiveness and ticket deflection. Validated content enables faster resolution and better self-service outcomes.
• Request history: Sera AI leverages request histories out of the box to power classification and similar-request detection. Organizations with more historical data will see better classification accuracy.
• Keyword optimization: Using the Suggest Improvements feature on knowledge articles helps optimize keywords, which improves search results for both the Virtual Agent and traditional search.

‍