agent-miner

Mining agent interactions in ServiceNow surfaces repetitive patterns, common resolutions, and handoff points from virtual agent transcripts and live conversations. It generates prioritized automation candidates with ROI estimates for bot implementation.

Calculate ROI for virtual agent automation candidates by analyzing live agent transcripts and incident workflows to detect high-frequency resolution patterns. The Skill outputs prioritized automation candidate reports containing estimated ROI for each identified bot opportunity.

Analyzing agent handoff points requires access to interaction, sys_cs_conversation, sys_cs_message, incident, sc_req_item, task, sys_user, and kb_knowledge tables. The Skill uses MCP and REST interfaces to query these records and map virtual to live agent transitions.

Yes, automation mining works with live agent transcripts and incident workflows. It analyzes these records alongside virtual agent conversations to detect repetitive resolution patterns, identify handoff points, and surface gaps in knowledge coverage.

Finding knowledge coverage gaps involves mining virtual agent conversations and live agent transcripts for repetitive unresolved patterns. The Skill identifies missing knowledge articles and outputs a prioritized automation candidate report to bridge those coverage gaps.

Generate a prioritized automation candidate report by mining incident workflows and agent interactions for repetitive patterns and common resolutions. The Skill outputs a structured report containing automation candidates ranked by ROI estimates and frequency.