What is the AI product you will use when design a solution?

[major li]

Hi community,

May I know in your daily work, will you use any AI product to help with your solution design?

Any practise to share? Thank you. 

[Tanushree Maiti]

Hi @major li 

AS per my understanding, 

NowAssist : Now Assist is recommended to improve agent productivity by automating, summarizing, and accelerating resolution workflows. Key use cases include auto-summarizing chat conversations at closure, generating concise resolution notes from case history, and summarizing incidents for quick context, saving significant manual documentation time.

Key Scenarios for Now Assist Recommendations

• Chat Summarization: Recommended when a virtual agent transfers to a live agent, when a live agent takes over from another, or when closing a chat conversation. It reduces wrap-up time by producing a condensed version of the conversation.
• Resolution Summarization: Recommended for generating final notes that reflect full context from cases, incidents, or chat transcripts, ensuring consistent documentation and better reporting.
• Incident/Case Summarization: Recommended when an agent needs to quickly understand the context of a long-running incident, case, or interaction in Agent Workspace.
• Ad Hoc Assistance: The Now Assist panel can be used to generate summaries directly within the Agent Workspace for CSM or ITSM, improving productivity and reducing manual data entry. 

Agentic AI: It is is recommended when you need to move beyond simple generative AI (summarization/chat) to autonomous, end-to-end task execution that bridges insights with actions

Key Scenarios for Recommending Agentic AI:

• High-Volume, Multi-Step Tasks: When a task requires gathering data, making decisions, and updating records across IT, HR, or Customer Service (e.g., employee onboarding/offboarding, complex incident resolution).
• Proactive Service Delivery: When AI can analyze data patterns to initiate actions before a human submits a request (e.g., updating systems based on upcoming role changes).
• Automation of Legacy Systems: When needing to update systems lacking modern APIs (e.g., HR, Payroll, or 20-year-old internal systems).
• Reducing "Un-automatable" Work: Handling exceptions and chaotic, non-scripted processes that traditional, rigid Workflow Designer flows cannot handle.
• Intelligent Routing and Triage: Using agents to analyze, prioritize, and route requests to reduce manual triage efforts

Predictive Intelligence: ServiceNow Predictive Intelligence (PI) is recommended when organizations need to automate high-volume tasks (e.g., >10,000 monthly tickets), reduce manual triage, and improve accuracy in categorization and assignment using historical data. It is best applied for streamlining service desk operations, accelerating resolution times, and reducing human error in ITSM, HR, and Customer Service workflows.

Key Scenarios for PI Recommendations

• High-Volume Triage: When analyzing large volumes of incoming incidents or cases to automate routing decisions (>10,000 monthly).
• Auto-Categorization & Assignment: When you have stable category structures and large historical datasets (>30,000 records) to train models for automatic field population (category, assignment group, priority).
• Similarity Detection: To surface relevant past resolutions for agents, reducing MTTR (Mean Time To Resolution) and providing proactive support.
• SLA Breach Prediction: To predict potential breaches and trigger automated escalations for high-priority tasks.
• Pattern Recognition: To group similar, ongoing incidents (clustering) to identify potential major incidents before they impact users