The Platform / Domains / AI Workforce Operations
Every employee already has an HR record. Every AI agent needs one too. XOPS maintains both on a single operational spine.
Two workforces
One org spine, one control plane
Continuous
Governed as reality changes, not reconciled after
Weeks, not quarters
To the first production Outcome
The category shift
Every previous technology wave gave employees better tools.
AI created an entirely new workforce.
XOPS is the control plane for AI Workforce Operations.
XOPS maintains the live AI Position of every employee and every agent: the same ownership, budget, policy, autonomy, and lifecycle primitives, applied identically to both workforces and attributed to one org spine.
Employees
Agents
Persona
Identity
Manager
Owner
Budget
Budget
Policy
Policy
Autonomy
Autonomy
Lifecycle
Lifecycle
Declared outcome
Every seat and every agent stays inside its budget, governed by policy, and attributed to a real owner.
Employee AI Position
Persona · Models allowed · Tools allowed · Monthly budget · Autonomy ceiling · Warn / throttle / suspend thresholds
Agent Workforce Position
Identity · Owner · Cost center · Budget · Tokens · Agent Workforce Index · Autonomy ceiling
ONE ORG SPINE
Department · Cost center · Region
Governed action
Warn · Throttle · Suspend · Reassign · Reclaim · Fund
What changes
Four ways the same AI Position changes the operating model for both workforces at once.
1. Dashboards report. A Position governs.
A dashboard tells you what was spent. XOPS maintains the Position that determines what should happen next, for every seat and every agent.
2. Budgets are static. A Position enforces continuously.
Warn, throttle, and suspend thresholds are evaluated continuously against actual consumption, not reconciled at the end of the month.
3. Agents are invisible. A Position attributes every one.
Every agent and every API key resolves to an owner, a cost center, and a workforce score. Unattributed spend becomes a visible governance gap, not a mystery line item.
4. Alerts arrive late. A Position acts before it compounds.
A runaway agent or an approaching budget is throttled or escalated the moment it crosses a threshold, not discovered when finance closes the books.
One operating model, not two.
Employees and agents look different on the org chart, but they are governed by the identical primitives: a persona or identity, an owner, a budget, a policy, an autonomy ceiling, and a lifecycle.
That symmetry is what makes both workforces governable from the same control plane.
Not a concept. A screen.
The Agent Consumption view inside XOPS Experience Center: every agent priced, scored, and attributed to an owner, live.
Pillar 02 · Agents · Last 30 Days
Agent AI consumption 1
Autonomous agents and API keys owned by cost center and region, scored by AWI, priced by tokens, flagged when unattributed.
Agent Spend
$187,305
Last 30 days
Active Agents
32
100% cost-center owned 2
Employee Pillar
Seat fit & rightsizing
$122,000 recoverable 3
Agent Pillar · AWI
Agent Workforce Index 4
17
Green
9
Amber
6
Red
Detail · Agents by Cost Center
Representative product view, XOPS Experience Center. Agent names, cost centers, and figures shown are illustrative.
The employee AI estate
A persona is an entitlement catalog, not a budget line. It defines which models an employee may use, which tools they may reach, what their monthly budget is, and how much autonomy their AI assistant is allowed to exercise on their behalf. Limits are defined once on the persona and enforced for every employee mapped to it.
Executive
Autonomy L3Models
Claude Opus · GPT-5.1
Tools
Claude for Work · ChatGPT Enterprise · Microsoft 365 Copilot
75%
Warn
90%
Throttle
Manager
Autonomy L2Models
Claude Sonnet · GPT-5.1 Mini · Copilot M365
Tools
Claude for Work · Microsoft 365 Copilot
Budget
$1,000/mo
80%
Warn
95%
Throttle
100%
Suspend
Security Analyst
Autonomy L2Models
Claude Sonnet · GPT-5.1
Tools
Claude for Work · ChatGPT Enterprise
70%
Warn
90%
Throttle
Finance Analyst
Autonomy L1Models
GPT-5.1 · Claude Sonnet
Tools
ChatGPT Enterprise · Claude for Work
Budget
$700/mo
80%
Warn
95%
Throttle
100%
Suspend
Illustrative persona configuration from a live product view. Models, tools, budgets, and thresholds are configured per organization.
Autonomy ceiling is not a model choice. It is a governance decision.
Everyone talks about which models an employee may use. Very few enterprises are talking about how much authority the AI acting on that employee's behalf may exercise without a human in the loop. The autonomy ceiling makes that decision explicit, per persona, and enforces it continuously.
The agent workforce
Every autonomous agent is continuously scored across four dimensions, then banded green, amber, or red. AWI is the single number that tells a leadership team whether the agent workforce is governed, not just running.
Efficiency
Value produced per dollar and per token spent.
Identity
Whether the agent resolves to a real owner and cost center.
Economic
Spend against budget, fit, and expected demand.
Fidelity
Whether the agent operates within its declared scope.
Autonomy ceiling
L1
Drafts and recommends. A human executes every action.
L2
Executes routine actions within scope. Exceptions route to a human.
L3
Executes broadly within policy. Reviewed continuously against AWI.
Priority actions
What XOPS surfaces automatically, ranked and routed to the owner, without a manual review cycle.
Support Triage Agent 020
Agent workforce risk: review governance, scope, and economic alignment.
Security Ops Agent 025
Agent workforce risk: review governance, scope, and economic alignment.
Security Ops Agent 015
Agent workforce risk: review governance, scope, and economic alignment.
Illustrative product view. Agent names and scores are representative of a live governance workspace.
Rightsizing
The same discipline that rightsizes software licenses and device fleets now applies to AI seats and agents, routing every over-provisioned entitlement to an owner for action.
Credential hygiene
API keys don't get offboarded when an employee leaves. XOPS finds every credential still active after its owner has departed.
Illustrative scenario
When the owner leaves
XOPS already knew Employee A owned Agent X, that it runs on API Key K, bills to Cost Center C, spends about 2.3M tokens a day against an $8,000/month budget, and reports to Manager B.
Without a Position
With XOPS
One departure. Every assistant, key, and agent accounted for.
Nothing left spending in the dark.
Coverage, not a pilot
The operating properties required to govern AI across a complex enterprise estate.
200+
Enterprise & AI integrations
4
Major AI providers covered
Continuous
Enforcement, not alerting after the fact
Above your stack
Deploys on the systems you already run
For CFOs, CIOs, CISOs & FinOps leaders
A 15-minute working session. No prep required.
Not another FinOps dashboard. The control plane for AI Workforce Operations.
Your next workforce doesn't report to HR.
It reports to XOPS.
