Virtual agents fail when they can't understand what customers are actually saying. Customer Service NLU Model for Virtual Agent Conversations brings pre-built natural language understanding models purpose-built for customer service interactions, enabling virtual agents to accurately interpret customer intent and respond meaningfully — without extensive training or configuration.
Built natively on ServiceNow, the NLU models integrate directly with Virtual Agent to improve intent recognition across common customer service scenarios. Teams get accurate, out-of-the-box language understanding tuned for CSM use cases, reducing the time and effort typically required to train custom NLU models from scratch.
Organizations deploying Virtual Agent see faster deflection of routine inquiries, reduced live agent escalations, and improved customer satisfaction.
- Pre-built CSM intent models: Purpose-trained NLU models for common customer service scenarios, ready to deploy.
- Virtual Agent native: Integrates directly with ServiceNow Virtual Agent — no third-party NLU service required.
- Reduced training overhead: Skip custom model training with models already tuned for customer service language.
- Improved deflection rates: More accurate intent recognition means fewer unnecessary escalations to live agents.
New
- Customer Service NLU Model for Virtual Agent Conversations is now available as a standalone store application. Auto-installs on instances upgraded to the Australia release or later.
Changed
- Delivery model updated from platform plugin to store app for independent versioning and faster updates.
Removed
- Previous plugin-based delivery is deprecated and will be removed in the C release.
- Virtual Agent — com.glide.cs.chatbot
- Customer Service Management (CSM) — com.sn_customerservice
- NLU platform — com.glide.nlu
- ServiceNow platform: Australia release or later