St. Luke’s Turns 400,000 Patient Interactions Into an Agentic AI Use Case

St. Luke's Hospital in Kansas City, Missouri has transformed 400,000 patient interactions into an agentic AI use case. Initially, the hospital faced a labor-intensive operation, with 100,000 labor hours spent monthly on tasks like prescription requests, appointment scheduling, and clinical consultations. These tasks were labor-intensive, costing the hospital over $3.5 million annually.

NextGen Coding proposed an agentic AI call center to handle routine workflows, with the system transcribing calls, classifying their intent, and routing them to either autonomous agents or human operators based on set escalation policies. This system also tracked automation rates, handle times, and escalation frequencies, with clinical oversight teams reviewing flagged interactions.

For human-assisted calls, the system could gather information before the handoff, giving clinicians useful context at the start of the consultation. The transformation was gradual, starting with conservative automation thresholds and staff onboarding, then expanding into scheduling and prescription workflows. After steady state, labor hours could drop from 400,000 to around 120,000 per month, with labor costs falling from about $3.57 million to approximately $965,536.

This would save the hospital over $2.6 million monthly and more than $31 million annually, accounting for AI infrastructure costs. The agentic AI model allows health systems to use human expertise where it's most needed while allowing software to handle a larger share of routine workloads.