The City of Raleigh is attempting to decouple municipal service capacity from headcount growth by deploying autonomous AI agents to manage internal workflows. By implementing ServiceNow’s IT Service Desk AI Specialist, the city aims to address the mounting pressure on its lean IT team, which currently consists of only four staff members supporting 4,400 employees. This deployment marks the first time a municipal government has moved ServiceNow’s autonomous worker technology into a live production environment to manage complex service desk tasks.
Scaling Municipal Services via Ral-E and Alli
Raleigh is managing rapid population growth by utilizing two distinct AI entities built on the ServiceNow AI Platform. The first, an agent named "Ral-E," functions as a guide for self-service requests across IT, HR, and facilities departments, achieving a 98% deflection rate by routing or answering queries via the city's knowledge base. The second, an L1 IT Service Desk AI Specialist named "Alli," moves beyond simple routing to resolve complete workflows. Alli utilizes historical incident data and remediation workflows to handle common requests from start to finish. This dual-agent approach has already resulted in 20+ automated workflows through ServiceNow’s Autonomous Workforce. Currently, nearly half of IT support requests are resolved autonomously, though the city has set a higher target of 85% autonomous resolution.
Operational Efficiency and Data Integrity
The integration of these AI specialists has significantly altered the city's cost structure and staff allocation. Raleigh reports a 66% reduction in IT service desk costs, a figure attributed to redirecting staff time away from routine connectivity and device questions. To maintain technical accuracy, help desk staff oversee Alli’s operations, ensuring the knowledge the agent acquires remains current. Data reliability appears high during this transition, with the city reporting that 98% of AI-generated incident summaries are accepted by staff without modification. Furthermore, the system has achieved a 98% correct routing rate for tickets on first contact. Looking ahead, the city plans to extend this model to its "Ask Raleigh" resident portal, potentially providing service to nearly 500,000 residents.
Key Takeaways
- Raleigh achieved a 66% reduction in IT service desk costs by redirecting staff time through AI automation.
- The city utilizes two AI entities: "Ral-E" for guidance/routing and "Alli" for end-to-end workflow resolution.
- Current autonomous resolution sits at nearly 50%, with a strategic target of 85% resolution.
TechInsyte's Take
In our view, Raleigh’s deployment serves as a critical proof of concept for the "agentic enterprise" within the public sector. By moving from simple chatbots to "specialists" that execute remediation workflows, the city is testing whether autonomous agents can handle the high-stakes governance required by government agencies. The 66% cost reduction is a significant metric, but the real strategic value lies in the 98% acceptance rate of AI summaries, which suggests that the technology is meeting enterprise-grade accuracy standards. If Raleigh successfully scales this to its 500,000 residents, it will demonstrate that AI can bridge the gap between shrinking municipal budgets and expanding citizen expectations.
Questions & Answers
How does the City of Raleigh plan to scale this AI deployment?
The city intends to extend the ServiceNow model beyond internal operations to the "Ask Raleigh" resident portal, aiming to provide service to nearly 500,000 residents via automated answers and service requests.
What specific metrics indicate the success of the AI agents?
Key performance indicators include a 66% reduction in IT service desk costs, a 98% deflection rate for Ral-E, and a 98% accuracy rate for both ticket routing and AI-generated incident summaries.
What is the distinction between the Ral-E and Alli AI agents?
Ral-E acts as a guide that answers questions and routes employees to the correct teams, whereas Alli is an L1 IT Service Desk AI Specialist designed to resolve complete workflows from start to finish.
How is the city ensuring the accuracy of the autonomous workflows?
Help desk staff actively oversee Alli’s work and are responsible for keeping the knowledge the agent learns accurate and current.
Source: ServiceNow
