Version history

Version 8.0.5 - March 2026

• New:
  • Voice Agent Evaluation
    • Evaluate ServiceNow voice agents at scale using transcripts and execution plans from historical runs.
    • Two default voice-specific evaluation judges: Task Completeness and Tool Choice.
    • Unified agent selection and results view within existing Agentic Evaluation workflows.
  • Enhanced User Experience – Agentic Evaluations
    • Streamlined onboarding to reduce the learning curve for new users.
    • Simplified setup workflows to accelerate evaluation configuration for all users.
  • Third-Party Model Support for Agentic Evaluation
    • Customers can select external models to power automated evaluations.
    • Enables alignment of judging behavior with organizational trust standards, preferences, and use-case requirements.
  • Model Versioning Enhancements
    • Structured upgrade process for model transitions.
    • Deprecation and retirement of older model versions instead of forced migrations.
  • AICT Integrations for BYOLLM Approval Flows
    • Formal approval flows for custom models.
    • Improved oversight and compliance for enterprise AI deployments.
  • Granular Admin Roles
    • The legacy broad Admin role is split into smaller, task-specific roles.
    • Introduction of a BYOLLM-specific admin role.
  • Structured Output Support
    • AI responses can be constrained to predefined JSON schemas.
    • Enables consistent parsing in agentic workflows and reduces ambiguity in downstream automation.

Version 1.0.2 - August 2024

Now Assist Skill Kit is a skill authoring tool designed for GenAI app developers. Now Assist Skill Kit empowers users to create custom skills that are not available in the standard out-of-the-box catalog or by configuring them. It provides a streamlined interface to a subset of advanced AI tools, enabling users to build sophisticated skills effortlessly.