Top 10 AI Prompts and Use Cases and in the Government Industry in Durham

Durham County's county-wide AI rollout shows why AI matters for local government in North Carolina: the Moveworks-powered “DCoBot” was deployed across 30+ departments supporting 2,200+ employees to automate routine IT requests, delivering instant answers within the first 30 days and freeing service-desk staff for higher-value projects - a practical efficiency gain that reduces service cost and speeds employee productivity (Durham County Moveworks case study on IT automation).

At the same time, public-safety research from the National Institute of Justice highlights AI uses - from gunshot detection to video analytics and crime-forecasting - that can improve response and resource allocation (NIJ article on AI for criminal justice and public safety).

Successful, equitable adoption depends on governance: North Carolina's seven Principles for Responsible Use of AI in North Carolina stress human oversight, transparency, and data quality - and local staff can gain those practical skills through training such as Nucamp AI Essentials for Work bootcamp (15-week).

"This transformation is not just about technology; it's about setting a new standard for public service in the digital age." - Chasity Locke, Assistant Director of IT Engagement and Communications

Selection prioritized real-world impact, regulatory alignment, and low-risk deployment: each candidate use case had to demonstrate measurable operational benefit (for example, Durham County's Moveworks “DCoBot” handled large volumes of routine IT requests across 30+ departments supporting 2,200+ employees within the first 30 days), comply with North Carolina's AI governance principles, and be assessable through a Privacy Threshold Analysis or the NIST AI Risk Management Framework; see the state's North Carolina AI Framework for Responsible Use and guidance on North Carolina AI Assessments guidance.

Practicality also mattered: solutions were scored for integration ease with existing stacks (ServiceNow/Teams), need for workforce training, and exposure to data risks - drawing on NCDIT guidance about publicly available generative AI and documentation practices.

The result is a top-10 list that favors pilots that can be risk-assessed quickly, produce clear time-or-cost savings, and scale responsibly across county departments.

"We're working on adoption - it's ramping up, and people are beginning to use it. Has every department embraced it fully? Not yet. But we're starting to shift the culture," - Zawadi Powell, IT Project Manager

Durham County's Moveworks deployment - branded DCoBot - puts autonomous IT support where employees already work, integrating Moveworks Copilot with Microsoft Teams and the county's ServiceNow ITSM to serve as the first point of contact for routine requests across 30+ departments and 2,200+ staff; within the first 30 days the bot handled a substantial number of routine inquiries, freeing service-desk teams for complex incidents and strategic projects and delivering immediate, 24/7 answers that remove bottlenecks and boost productivity (Durham County Moveworks deployment case study).

Implementation prioritized a growing knowledge base and tight integrations, and the county continues to focus on adoption and trust-building while planning expansions (for example, HR topics) that could further cut administrative load; technical teams can replicate this pattern using pre-built connectors such as the Moveworks Microsoft Teams integration for faster deployment to speed deployment and user discoverability.

"We're getting to a point where the bot is starting to do more of the transactional work, that includes providing step-by-step instructions to common IT inquiries or serving up forms for easy access and completion." - Zawadi Powell, IT Project Manager

Building on the DCoBot pattern, a Durham County HR Bot would integrate with existing Microsoft Teams and ServiceNow workflows to deliver instant, self‑service answers to high‑volume HR questions (benefits, leave, basic policy), surface fillable forms for onboarding and paperwork, and escalate sensitive or complex cases to human specialists - freeing HR staff from routine triage and speeding employee access to services.

A focused pilot uses the same tight integrations and a growing knowledge base to limit scope, validate accuracy, and protect data; practical steps for local government pilots and playbooks can guide that rollout (Durham government AI pilot practical steps and playbook), while clear KPIs (usage, MTTR, and cost‑savings) measure impact and guard ROI (AI metrics for Durham government: usage, MTTR, and cost‑savings).

Plan staffing and maintenance around roles that sustain integrations and governance - such as an Applications Development Analyst - to keep the bot reliable and compliant (Applications Development Analyst job description and responsibilities).

AI-powered ticket triage turns noisy queues into prioritized, actionable work by intercepting new tickets, using a customer-scoped index of historical tickets to predict fields (assignment group, category, cmdb_ci), and - when confidence thresholds are met - updating records automatically within roughly 30 seconds to one minute; in practice that means Durham's DCoBot pattern can deflect routine L1 work into autonomous resolution or precise routing so human teams see fewer misclassified handoffs and faster incident handovers.

The approach depends on good historical data and configurable confidence thresholds, and it scales: Moveworks' triage and agentic systems show enterprise benefits such as dramatically lower MTTR (Leidos reported a reduction from 47 hours to 15 minutes, a ~180x improvement), while enabling secure integrations with ServiceNow and Microsoft Teams for county deployments.

The practical payoff: fewer ticket reassignments, faster specialist response, and immediate cost savings on recurring, high-volume requests. Metrics - Coverage: Proportion of tickets triaged automatically (example: 300/1,000 = 30%); Precision: Correct predictions among triaged tickets (example: 150/300 = 50%).

North Carolina's rapid statewide rollout of AI-based traffic signal software - now deployed at roughly 2,500 intersections - offers a concrete blueprint for Durham to pilot corridor-level signal optimization that prioritizes safety and congestion mitigation; Flow Labs' machine‑learning platform reportedly fuses connected‑vehicle feeds and existing detection data to provide continuous intersection monitoring and data‑driven timing adjustments, while state guidance highlights traffic signal management as a high‑value AI use case for local governments (StateScoop report on North Carolina AI traffic signal deployment, Safety21 analysis of Flow Labs connected-vehicle modeling, ncIMPACT overview of AI uses in North Carolina government).

A complementary NCDOT multimodal pilot around NCSU - 27 upgraded intersections, new roadside controllers, and the YU2X app backed by a $2M ATCMTD grant - demonstrates how paired infrastructure and mobile feedback can surface real‑time signal phases and pedestrian alerts; the practical takeaway for Durham: start with a 1–2 corridor pilot using existing controller upgrades and connected‑data feeds to measure delay reduction and pedestrian safety improvements before scaling.

“North Carolina didn't just invest in a new tool - they embraced a new model for traffic operations.” - Jatish Patel, CEO of Flow Labs

Durham County Tax & Permits can cut processing time, reduce appeals, and make valuations more objective by combining Wake County's machine‑learning approach to mass appraisal with Chatham County's GIS-driven permitting and appeals tools: Wake's SAS‑powered models process roughly 400,000 properties and ingest thousands of daily sales to support a move from an eight‑year to a four‑year reappraisal cadence, producing objective, repeatable estimates that limit time‑consuming appeals (Wake County machine-learning appraisal program and results), while Chatham's use of ArcGIS Survey123, an online appeals form and a comparable‑sales mapping app streamlined appeals management and collected over 100,000 field photos to validate parcel records and speed reviewer decisions (Chatham County GIS-enabled appraisal and online appeals workflow).

The practical payoff for Durham: fewer manual inspections, faster permit clearances, clearer comparable sales for residents, and more assessor time for complex cases - measurable wins in cost, transparency, and trust that scale across county departments.

"We needed unbiased support to analyze our volumes of data, and SAS was the obvious choice." - Marcus Kinrade, Revenue Director, Wake County

Public-safety analytics such as gunshot detection and incident triage offer Durham a focused way to speed verified alerts to dispatchers and improve resource allocation, but they require constrained pilots, human-in-the-loop checks, and strict measurement to avoid costly false positives; start small - instrument one patrol beat, route sensor alerts into an auditable triage workflow, and require dispatcher confirmation before dispatch - and use statewide playbooks to design scope and oversight.

Referenceable resources for that approach include practical steps for local-government AI pilots (practical steps for local-government AI pilot projects in Durham) and the GovAI Toolkit and North Carolina pilot examples (GovAI Toolkit and North Carolina pilot examples for government AI); prioritize clear KPIs - MTTR, usage, false‑positive rate, and cost per verified incident - to prove operational value and meet county governance standards (metrics to measure AI impact in Durham government operations).

The payoff for Durham is a measurable, auditable improvement in response prioritization that informs whether to scale safely across neighborhoods.

ncIMPACT's applied analytics turn statewide evidence into actionable options that Durham leaders can use to set priorities, allocate limited housing dollars, and design measurable pilots: its housing analysis flags a statewide shortage of 207,837 affordable rental units and finds 72% of extremely low‑income households spend more than 50% of income on housing, a concrete scale-of‑need metric that signals why targeted rental production and preservation matter now (ncIMPACT analysis of North Carolina's housing shortage).

The initiative's practice-oriented stance - providing neutral data, short/medium/long‑term policy options, and county‑level toolkits - helps translate those statistics into municipal need analyses, regional housing plans, and workforce‑support strategies such as teacher housing pilots (ncIMPACT Initiative homepage and resources).

Its companion work on AI use cases further shows how analytics can be operationalized (for example, traffic signal optimization or automated appraisal support) while emphasizing governance, data standards, and pilot evaluation that Durham can adopt to ensure policy choices are both evidence‑based and auditable (ncIMPACT report on AI uses in North Carolina).

The practical payoff: clear baseline metrics to measure progress, reduce appeals and inefficiencies, and build consensus across stakeholders before scaling investments.

A Durham Public Records Bot should make it faster for residents to find and request government documents while protecting legally sensitive information: integrate the bot with the City of Durham NextRequest public records portal so it can surface already-posted records, suggest targeted search terms, and guide users to create clearer requests (the City portal shows 8,438 requests and counting, a concrete signal of demand for easier discovery) - pair that front-end convenience with clear guardrails from the North Carolina Department of Information Technology (NCDIT): never submit PII or confidential data to publicly available generative AI tools and treat prompts/outputs as potentially public, and require reviewers to log and verify any AI-assisted responses - finally, design the bot so every retrieved or generated record entry conforms to North Carolina's archival standards for trustworthy digital public records (indexed, timestamped, and retention-tagged) to preserve admissibility and auditability per State Archives guidance.

Key references: City of Durham NextRequest public records portal (City of Durham NextRequest public records portal - public records requests and search), NCDIT guidance on using publicly available generative AI (NCDIT guidance - do not submit PII to public generative AI), and Guidelines for Managing Trustworthy Digital Public Records from the North Carolina State Archives (State Archives guidelines - managing trustworthy digital public records).

The practical payoff: faster citizen access, fewer repetitive help-desk requests, and a documented trail that meets state law - demand signal: City NextRequest shows 8,438 requests; NCDIT rule: do not enter PII into publicly available generative AI; records standard: State Archives requires trustworthy digital records with retention and admissibility controls.

Document-drafting assistance can cut routine drafting time for Durham staff - standard letters, permit notices, and policy summaries - while improving consistency, but only when paired with North Carolina's risk controls: use state‑procured or NCDIT‑approved tools (and state email for registrations), never paste PII or confidential case details into publicly available models, disable chat history for high‑risk prompts, and log tool name, model/version and prompts when content is substantially AI‑generated so outputs remain auditable and defensible; see NCDIT guidance on using publicly available generative AI and the state's AI Framework for Responsible Use for required safeguards (NCDIT guidance on using publicly available generative AI, NCDIT AI Framework for Responsible Use).

Complement these controls with UNC/academic disclosure practices - document tool versions, prompts, and post‑generation edits - and always treat AI output as a draft requiring independent fact‑checking before publication (UNC guidance on generative AI use in research); the practical payoff is faster, more consistent public communications that remain compliant with public‑records and privacy rules while preserving human review and legal accountability.

“The public has to put a certain amount of trust in the government,” - Cherie Givens, NCDIT Chief Privacy Officer

Durham can tap the Research Triangle's coordinated training ecosystem to staff and sustain safe, practical AI adoption: NC State's AI Academy trains entry‑level AI professionals across sectors and reports over 2,000 graduates, 100+ industry partners, and a $6M Department of Labor grant - a concrete pool of hireable, job‑ready talent for county pilots and retraining programs (NC State AI Academy training program).

AdvanceNC brings 18 colleges, universities, and workforce boards together to align curricula, apprenticeships, and employer pipelines across a 1.5 million labor force, reducing recruitment friction for local government AI roles (AdvanceNC talent pipeline coalition).

Regional college partnerships already win targeted funding - RTP Bio's $1.2M federal grant for biotech pathways is one example - showing how pooled resources accelerate upskilling and cut months off time‑to‑competency for municipal AI operators and auditors; the practical payoff is measurable: AI‑literate staff available within months to run, evaluate, and govern county pilots.

"The greatest strength of the AI Academy is the exceptional industry partnerships we have formed. Our consortium leads and informs the important work of the of building a pipeline of highly qualified and well-prepared AI talent for the U.S." - Carla C. Johnson, Ed.D., Executive Director and Principal Investigator, AI Academy

Durham's safest path to wide AI adoption is a pilot‑first roadmap that pairs the N.C. Department of Information Technology's living AI standards with fast, measurable pilots: run a Privacy Threshold Analysis and a NIST‑aligned assessment before deployment, limit scope to one corridor/beat or one department, and track clear KPIs (MTTR, usage, cost‑savings) while keeping humans in the loop - following the NCDIT AI Framework for Responsible Use and the state's NCDIT AI Assessments guidance.

Pair governance with workforce readiness: a 15‑week practical training like Nucamp's AI Essentials for Work teaches nontechnical staff to write prompts, use AI tools responsibly, and sustain governance - so Durham can prove ROI on a single, auditable pilot before scaling across departments and preserve privacy, transparency, and citizen trust.