ODOE AI Service Tier Catalog

AI Service Tier Catalog

Governed view of the AI tiers ODOE IT can use across self-service, intake, operations, and leadership support without losing budget discipline, auditability, or human accountability.

AI Governance Layer

Match AI capability to business risk, service need, and cost discipline.

This catalog defines which AI tier is appropriate for common ODOE use cases, what each tier costs in relative terms, where it can safely operate, and when work must escalate to a more capable model or back to a human owner. The point is not to use the most advanced model every time. The point is to use the right level of intelligence for the right decision.

Audience: IT leadership, service owners, governance Use: policy + budget + routing design Connected to: Request + Knowledge + Runbook + FinOps + Optimization
coverage
87%

Current AI-supported request volume addressable by standard tiers

budget guardrail
3

Tiers allowed for routine request handling

human owned
100%

Risk-bearing actions requiring accountable human approval

specialist use
6%

Work expected to justify premium reasoning tiers

AI Tier Definitions

Each tier is defined by relative cost, reasoning depth, operational speed, acceptable use, and governance expectations. The tiers are intentionally service-oriented rather than vendor-branded.

Policy-ready structure
Tier Cost / Speed Profile Best Fit Guardrail
Tier 1: Lightweight Assist
Lowest-cost, fast-response assistance for simple, low-risk interactions		 Lowest cost, highest throughput, shallow reasoning Knowledge lookups, form guidance, standard FAQ answers, request-field suggestions No policy interpretation, no priority changes, no unsourced answers
Tier 2: Standard Service Copilot
Balanced tier for everyday request shaping and summarization		 Low-to-moderate cost, fast response, moderate reasoning depth Request clarification, duplicate detection, triage summaries, draft status notes May recommend but not decide routing, severity, or approvals
Tier 3: Governed Analyst
Higher-confidence reasoning with approved enterprise context		 Moderate cost, slightly slower, stronger synthesis and comparison Runbook step preparation, evidence comparison, problem clustering, change risk narratives Requires approved sources, write-back logging, and human review before external impact
Tier 4: Specialist Operations Advisor
Advanced reasoning for higher-risk operational and cross-system scenarios		 High cost, slower, deeper multi-step reasoning and expert-level analysis Major incident support, vendor escalation packets, dependency-aware impact assessment, executive brief drafting Use only for higher-severity work, material service risk, or leadership-level synthesis
Tier 5: Restricted Premium Specialist
Premium tier reserved for rare, high-value, high-governance situations		 Highest cost, lowest routine volume, deepest domain and synthesis capability Cross-portfolio scenario analysis, complex root-cause synthesis, high-stakes policy support, sensitive multi-source reasoning Formal approval required, restricted datasets only, no autonomous action of any kind

How The Platform Chooses A Tier

Routing should not depend only on user preference. It should reflect service type, urgency, risk, data sensitivity, and whether the output will affect people, systems, or public-facing information.

Severity

Incident impact

Higher-severity incidents move upward from routine assistance to governed operational reasoning.

Category

Work class

Knowledge, intake, incident response, approvals, and leadership reporting should not use the same AI tier by default.

Risk

Decision exposure

If the output can affect service, access, policy, or public information, the routing model must shift to stricter guardrails.

Data

Sensitivity

Restricted sources and regulated data narrow both the eligible tier and the approved source set.

Budget

Cost efficiency

When work is routine and low-risk, the system should route down to the least expensive safe tier.

Recommended Tier By Service Pattern

This is the operating translation layer between platform services and AI cost / capability choices.

Low-Risk Service Work

Best for guided self-service and intake assistance where speed and budget matter more than deep reasoning.

Tier 1 knowledge prompts Tier 2 request shaping human escalation on ambiguity

Governed Operations Work

Best for incident evidence packs, problem review summaries, and change-risk synthesis with approved data sources.

Tier 3 primary Tier 4 for severity 1-2 audit write-back

Leadership And Sensitive Analysis

Best for cross-system impact, executive scenario framing, and premium reasoning where low confidence is unacceptable.

Tier 4 briefing support Tier 5 rare specialist use formal approval required

Tier Comparison

Capability increases do not replace governance. Higher tiers are valuable only when the problem justifies the cost and control model.

Speed

Tier 1 and Tier 2 optimize for rapid interaction at scale. Tier 4 and Tier 5 trade speed for deeper, more careful synthesis.

Cost

Routine work should stay in lower tiers by default. Premium tiers should be justified by risk, complexity, or decision value.

Reasoning depth

Multi-source comparison, dependency analysis, and incident command support belong in governed higher tiers, not lightweight chat assistance.

Human accountability

Every tier can assist. No tier can approve access, publish policy, execute service-impacting changes, or close incidents autonomously.

A mature AI platform does not start with the most powerful model. It starts with explicit service tiers, clear guardrails, and a routing strategy that respects cost, risk, and human accountability.