Grand Traverse County Goes Cloud First with AI Pilot and Justice Center Planning

Grand Traverse County’s commissioners are poised to approve a coordinated set of technology and facilities moves that together mark a deliberate shift toward cloud-first operations and early, managed adoption of generative AI — actions driven by last year’s ransomware disruption, practical vendor roadmaps, and an appetite for operational resilience that comes at a clear and recurring fiscal cost.

Background​

Grand Traverse County’s agenda packages and public reporting show three tightly linked priorities: migrate major enterprise systems to cloud-hosted platforms, subscribe to Microsoft 365 with a measured rollout of Microsoft Copilot seats to create an internal AI pilot cohort and a nascent “Center of Excellence in Artificial Intelligence,” and contract an external consultant to plan a consolidated justice center (including a new jail) through a formal RFP. These actions follow a disruptive ransomware incident that accelerated cloud moves for mission-critical public-safety systems and prompted a broader review of security and continuity strategies.
The county’s official paperwork lists specific line-item estimates for the next 12–24 months and frames the debate as a trade-off between higher recurring subscription fees and significantly improved operational resilience, security patching cadence, and vendor-managed upgrades. Those budget numbers are drawn from the county meeting packet and public reporting; they may shift during contract negotiation and procurement.

---

What commissioners will decide today​

• Approve a cloud upgrade for the county’s Enterprise Permitting & Licensing (EPL) system (used by Construction Codes, Environmental Health, and Equalization/GIS), with an estimated first‑year cost of about $274,603 and an annual recurring cost near $256,003 thereafter — compared with the current EPL subscription of just over $82,000.
• Authorize Microsoft 365 subscription renewals totaling nearly $400,000, a package that includes 100 Microsoft Copilot licenses intended to seed a pilot cohort across county offices — a $36,000 increase over last year’s licensing spend. The county cites this step as the beginning of AI‑led productivity experiments and staff training, and as the foundation for its proposed internal Center of Excellence.
• Approve a formal RFP process to hire a consultant to conduct feasibility, pre‑design, and (optionally) construction‑oversight services for a proposed justice center, with multiple siting and renovation/building scenarios to be evaluated.
• Hear the county treasurer’s presentation after an audit found a clean opinion but flagged timeliness issues with bank reconciliations and recommended regular reporting on investments.

Each of these discrete votes is operationally and politically connected: cloud migration responds to cyber risk and maintenance burden; AI licensing responds to productivity and administrative priorities; and justice‑center planning sets the capital and fiscal backdrop for multi‑decade debt decisions.

---

Technology upgrades: from on‑prem to cloud — what’s changing​

Why the county is moving​

Grand Traverse County moved its public‑safety computer‑aided dispatch (CAD) to cloud hosting after a ransomware attack disrupted services, and District Court JIS has also moved. County IT leadership frames cloud adoption as a security and continuity imperative — cloud models deliver rapid security patching, managed backups, and geographic redundancy so data and services remain accessible even if one site is impacted. Those arguments reflect common post‑incident recovery playbooks used by many local governments.

The EPL upgrade: capabilities and costs​

Tyler Technologies’ Enterprise Permitting & Licensing (EPL) is the vendor product the county proposes to migrate to a cloud‑hosted model. Tyler publicly markets EPL as a cloud‑capable permitting and inspections suite with GIS integration, mobile inspection apps, and online submittal features — a functional match for county permitting workflows. Migrating EPL to vendor‑hosted cloud services will deliver 24/7 public portals, centralized workflows, and vendor‑driven security updates.
The county’s stated estimate — a first‑year outlay of roughly $274,603 and an annual recurring cost of about $256,003 (with 2–5% annual increases expected) — represents a substantial uplift versus the current EPL subscription of ~ $82,000. That delta captures hosting, migration services, and a higher support tier, but it also converts capital and local operations burden into vendor OPEX. The county packet provides these figures as planning estimates; final contract terms, credits, or phased payment options could alter the realized cost.

What cloud buys (and what it doesn’t)​

Cloud hosting gives the county:

• Continuous security patching and managed backups, reducing the “single-site” failure risk.
• Automatic software upgrades, minimizing the need for local compatibility testing for minor releases.
• Geographic redundancy so critical records survive localized physical incidents.

Cloud hosting does not automatically solve:

• Integration complexity (GIS, court systems, tax, financial systems still require connectors).
• Vendor lock‑in unless contracts include clear egress, data portability and escrow provisions.
• Egress and long‑term OPEX budgeting — recurring fees can be sticky and grow across multiple systems.

These caveats are material: cloud migration shifts operational responsibility but introduces contract and long‑term licensing exposure that must be modeled in five‑ to ten‑year budgets.

---

Artificial intelligence: Copilot, Center of Excellence, and pilots​

What the county plans to buy and why​

The county’s renewal includes 100 Microsoft 365 Copilot licences as part of a near‑$400,000 Microsoft subscription package. The IT director frames Copilot as a way to introduce “AI‑powered productivity capabilities” and to train a distributed cohort of users across nearly 30 offices who will trial Copilot and provide feedback. The administration expects the pilot to be accompanied by formal training and a future county AI‑use policy.

Pricing verified​

Microsoft lists Microsoft 365 Copilot for enterprise at roughly $30 per user per month (annual commitment) and business bundles that include Copilot start at price points above that. That public pricing aligns with the county’s budget math: 100 seats at $30/month works out to $36,000 per year — which matches the reported licensing increase in the county packet. This consistency strengthens confidence that the county’s licensing projection is grounded in widely published vendor pricing.

Where Copilot makes sense — and where it’s risky​

Copilot can deliver fast wins in administrative tasks, for example:

• Drafting standard correspondence, memos, and template letters.
• Summarizing long documents (subject to verification).
• Assisting with data analysis inside Excel via natural‑language queries.

But the hazards are real:

• Hallucination risk: generative models can fabricate facts or misstate citations; human verification remains mandatory for legal, fiscal, and public‑safety outputs.
• Data governance: connecting enterprise data sources to Copilot requires careful controls to prevent inadvertent disclosure or retention of sensitive data.
• Operational cost scale: a pilot of 100 seats is affordable; broadening to hundreds of staff would multiply recurring costs and require sustained budget allocation.

The county’s plan to pair licenses with training and a governance Center of Excellence is a best practice — provided governance enforces human‑in‑the‑loop checks, audit logging, and clear approval gates before AI outputs are relied upon for decision‑making.

---

Public safety pilots: AI in dispatch​

Grand Traverse County already deployed an AI call‑handling pilot for non‑emergency Central Dispatch calls to triage and process lower‑risk traffic — a move designed to free dispatchers for emergency incidents. The county’s Central Dispatch also purchased a third‑party AI call‑handling product (Aurelian AI) to manage non‑emergency calls under a pilot contract; local reporting indicates the trial year cost was approximately $60,000 with follow‑on annual costs near $72,000 if retained. Commissioners discussed concerns about preserving the human touch for sensitive callers and retaining explicit opt‑outs to speak with human operators.
That pilot is illustrative: AI can reduce workload and response time, but any automation at the intake point for public‑safety calls requires tight KPI measurement, fallback triggers to human operators, and an incident‑logging trail for auditability.

---

Facilities RFP: planning the justice center​

Scope and phases​

The RFP commissioners will initiate requests a consultant to lead:

• Pre‑design and feasibility — building needs assessment, scenario modeling (renovation vs. new build), preliminary budgets and timelines.
• Design services — coordination with architects/engineers, permitting strategy and approvals.
• Optional construction oversight — subject to future commission approval.

The consultant’s task will be to evaluate options across multiple county campuses (Boardman, LaFranier) and the existing Law Enforcement Center, and to quantify lifecycle operating costs and funding scenarios.

Fiscal framing​

Financial advisers presented debt scenarios showing a manageable debt burden when capital costs are spread across decades. That framing is routine for large municipal projects, but it depends crucially on conservative contingency, interest‑rate assumptions, and alignment with the county’s capital plan. Commissioners should insist on transparent sensitivity analyses (interest‑rate stress tests and construction‑inflation contingencies) in consultant proposals to avoid mid‑project fiscal strain.

---

Treasurer’s audit presentation: governance issues flagged​

The county’s most recent audit delivered a clean opinion but was filed late with state authorities. The audit noted that bank reconciliations were not completed timely or accurately and recommended they be performed within 30 days. Auditors also recommended regular investment reporting to the commission. Treasurer Jamie Callahan supplied a monthly interest‑income chart through August 2025 and is expected to address the audit findings in the commissioner meeting. These governance recommendations are straightforward operational controls that should be remedied quickly to maintain fiscal transparency.

---

Technical verification and cross‑checks​

• The county’s EPL cloud migration numbers and Copilot seat counts are documented in the county meeting packet and reported locally; those figures represent planning estimates rather than final executed contract values. Readers and procurement officials should treat them as vendor‑negotiable until contracts are signed.
• Microsoft’s public pricing pages list Microsoft 365 Copilot at about $30/user/month for enterprise add‑on licenses, which aligns with the county’s stated $36,000 annual increase for 100 seats. This independent vendor pricing is a reliable cross‑check for the county’s budget math.
• Tyler Technologies’ published product pages confirm that Enterprise Permitting & Licensing is designed for cloud deployment and includes GIS, mobile inspection, and electronic plan review features — making Tyler a plausible cloud partner for the county’s permitting needs. That technical fit justifies the county’s consideration of an EPL cloud migration on capabilities grounds; the final contract should still be checked for SLAs, data portability, and compliance attestations.
• The decision to migrate CAD and certain court systems to cloud hosting after the 2024 ransomware event is corroborated by government‑technology reporting and local press coverage, which describe the incident’s operational impact and the subsequent vendor‑hosted recovery strategy. These independent accounts corroborate the county’s impetus for rapid cloud adoption for mission‑critical services.

---

Strengths of the county’s approach​

• Security-driven prioritization: The county prioritized mission‑critical public‑safety systems immediately after a ransomware incident and used that event to justify systematic improvements, which is a practical, risk‑based approach.
• Measured AI adoption: Buying a limited set of Copilot licenses for a pilot cohort, paired with training and a Center of Excellence plan, follows industry best practices for staged AI adoption and governance.
• Independent planning for capital projects: Hiring a consultant through an RFP to analyze justice‑center options (renovate vs. new build) will produce objective cost estimates and allow commissioners to evaluate trade‑offs before committing to long‑term debt.
• Transparency about recurring costs: The county’s public packet explicitly shows the recurring OPEX figures and expected annual increases, enabling more informed fiscal planning rather than burying subscription growth in vague line items.

---

Risks and weaknesses — what the commissioners should probe​

• Recurring cost escalation: Converting on‑prem costs into vendor OPEX concentrates long‑term financial risk in subscriptions that can escalate with seat counts, premium features, or unforeseen integration projects. Model a five‑year licensing trajectory before full rollouts.
• Contractual ambiguity and vendor lock‑in: The county must insist on clear data‑egress clauses, encryption key ownership/escrow, and termination terms to avoid being trapped with a vendor or incurring outsized migration costs if it later needs to change providers. Require SLAs for RTO/RPO, audit logs, and breach notification timelines.
• Integration and hidden professional‑services costs: Even with vendor hosting, integration work (GIS, BS&A, court systems, finance) often drives schedule slippage and extra professional‑services hours. Budget for integration, testing, and a dual‑run window.
• AI governance gaps: A Copilot pilot must be accompanied by enforceable policies on sensitive data usage, retention, and human verification. Without that, AI can create liability in legal, records, or public‑safety contexts. The county’s pledge to produce an AI policy is necessary but must be expedited.
• Public trust and human factors: Using AI in public‑safety intake or public‑facing permit decisions requires community outreach and clear opt‑outs. The dispatch pilot shows the sensitivity of public acceptance.

---

Practical recommendations and a procurement checklist​

To convert these strategic moves into durable, low‑risk outcomes, commissioners and procurement should require the following before final contract approvals:

• Contract and security checklist (must be included in solicitations and reviewed by county counsel and IT security):
• Explicit data portability and egress pricing language.
• Encryption key control statement (tenant keys, BYOK, or escrow).
• CJIS/FedRAMP/HIPAA attestations where applicable for public‑safety and court systems.
• Vendor incident response SLAs (time to notify, time to contain, and forensics support).
• Pricing escalation caps or formulae for long‑term renewals.
• AI pilot and governance gates:
• Define SMART KPIs for the Copilot pilot (e.g., time saved per permit, percent of non‑emergency calls handled by AI with human confirmation rates).
• Require human‑in‑the‑loop for any AI output that affects legal status or safety.
• Instrument audit logging and retention policies for AI interactions.
• Publish anonymized pilot metrics quarterly to maintain public confidence.
• Integration and cost modeling:
• Include independent cost estimates for integration, middleware, and five‑year OPEX totals in vendor proposals.
• Model worst‑case licensing scenarios (50–100% seat growth) and include those in budgetary forecasts.
• Budget for a dual‑run window and realistic egress charges during cutover.
• Justice center procurement standards:
• Require consultant proposals to include sensitivity testing for interest rate and construction‑inflation scenarios.
• Insist on transparent life‑cycle cost analysis (operating expenses for renovation vs. new build).
• Demand independent cost verification and public engagement plans as part of the scope.

---

Conclusion​

Grand Traverse County’s proposed mix of cloud migrations, AI pilot licensing, and structured facilities planning is a coherent response to a discrete operational shock and reflects a pragmatic modernization path: prioritize security and continuity, pilot AI through a controlled cohort, and seek outside expertise for major capital decisions. Those moves position the county to deliver more resilient services and potentially faster administrative workflows.
However, the county is exchanging one kind of complexity for another. The real work now is governance: writing enforceable cloud contracts with exit clauses, modeling multi‑year subscription costs, instrumenting AI pilots with stringent human oversight, and keeping the public informed as capital plans crystallize. If commissioners treat today’s votes as program authorizations rather than final solutions — and insist on measurable gates, independent verifications, and tight procurement guardrails — the investments can deliver measurable improvements without surrendering fiscal or operational control.
For municipal IT teams and administrations watching this case, Grand Traverse County is a live example of how a security incident can accelerate modernization — and why that momentum must be met with disciplined contracting, measurable pilots, and transparent governance to ensure new technology investments actually reduce risk rather than migrate it.

---

Source: The Ticker County Eyes AI/Tech Upgrades, Jail Planning Contract | The Ticker