Optimize AI privacy utility portfolio

What it's for

Creates a mathematically explicit privacy–utility frontier: which AI mechanisms preserve the most business value without overspending any shared population's formal budget or hiding correlated privacy downside.

What you give it

Inputs split into evidence read from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "implementation_budget": 2000,
  "privacy_accounts": [
    {
      "id": "customer-privacy-account",
      "maximum_epsilon": 3,
      "rdp_orders": [
        2,
        4,
        8,
        16
      ],
      "target_delta": 0.00001,
      "value_at_risk_scenarios": [
        50000,
        150000
      ]
    }
  ],
  "privacy_options": [
    {
      "attack_probability_scenarios": [
        0.2,
        0.3
      ],
      "dependency_option_ids": [],
      "direct_privacy_loss_scenarios": [
        1000,
        3000
      ],
      "evidence_verified": true,
      "exclusion_option_ids": [],
      "id": "support-current",
      "implementation_cost": 0,
      "is_current_state": true,
      "latency_ms_scenarios": [
        20,
        25
      ],
      "operating_cost_scenarios": [
        100,
        120
      ],
      "rdp_epsilon_by_order": [

What you get back

This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.

{
  "assumptions": [
    "Every option's RDP vector uses its account's exact order grid and already reflects the approved mechanism, clipping, sampling and neighboring relation.",
    "Utility, attack probability, direct loss, latency, operating cost and shared value at risk are prospective scenario-aligned effects from local validation, not labels inferred by an agent.",
    "Shared account compromise is priced once through joint survival plus a common failure state; direct option loss remains non-overlapping and separately governed."
  ],
  "decision": "adopt_governed_ai_privacy_utility_portfolio",
  "economics": {
    "expected_net_value": -868,
    "expected_net_value_gain_vs_current": 20756,
    "expected_privacy_loss": 9072,
    "implementation_cost": 1000,
    "privacy_loss_conditional_value_at_risk": 38600,
    "privacy_loss_value_at_risk": 38600,
    "risk_adjusted_score": -20168
  },
  "limitations": [
    "Exact mode proves optimality only for the supplied finite model; beam mode is feasible and deterministic but does not prove global optimality.",
    "Selection is not privacy certification, legal approval, deployment authority or a claim that epsilon alone measures expected harm or individual protection."
  ],
  "method": "multiple_choice_rdp_composition_privacy_utility_cvar_portfolio_v1",
  "pareto_frontier": [
    {
      "expected_net_value": -868,
      "implementation_cost": 1000,
      "option_ids": [
        "support-private"
      ],
      "privacy_loss_cvar": 38600
    },
    {
      "expected_net_value": -21624,
      "implementation_cost": 0,
      "option_ids": [
        "support-current"
      ],
      "privacy_loss_cvar": 114000
    }
  ],
  "privacy_account_usage": {
    "customer-privacy-account": {
      "best_rdp_order": 16,
      "composed_epsilon": 1.0675,
      "maximum_epsilon": 3,

How it works

Constrained optimization — Pick the best feasible option under real limits — budget, headcount, dependencies, capacity — rather than ranking a list and hoping it fits.

1. 1 Validate scenario-aligned utility, attack, direct loss, latency, cost and RDP vectors for every current and alternative workload mechanism.
2. 2 Compose selected RDP vectors at every account/order and reject portfolios exceeding target-delta epsilon, utility, latency, controls, relations, budget or resources.
3. 3 Price shared account compromise once through joint survival plus common failure, then solve exactly or by disclosed economic beam and expose the value-tail-cost Pareto frontier.

Before you trust it

Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.

Calibrating it to your org

Calibration workflow

1. 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
2. 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
3. 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
4. 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
5. 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
6. 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.

Call it from your AI

You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 18 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.

gitrevio_capabilities_search
  { "q": "select one validated privacy mechanism per" }
  → finds "optimize_ai_privacy_utility_portfolio"

gitrevio_capability_describe
  { "capability_id": "optimize_ai_privacy_utility_portfolio" }
  → returns the input schema and agent guidance shown on this page

gitrevio_capability_run
  { "capability_id": "optimize_ai_privacy_utility_portfolio", "arguments": { ... } }
  → returns the result shown above

Works in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.

Related tools

See every tool in AI risk, rights & assurance →