Estimate estimate at completion distribution

What it's for

Extends Gitrevio's probabilistic forecasting promise from task duration into finance-grade bottom-up p10/p50/p75/p90 EAC, budget-breach, and shared-shock ranges.

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.

{
  "budget_at_completion": 200,
  "components": [
    {
      "actual_cost_to_date": 10,
      "id": "backend",
      "remaining_cost_p50": 50,
      "remaining_cost_p90": 100
    },
    {
      "actual_cost_to_date": 10,
      "id": "migration",
      "remaining_cost_p50": 50,
      "remaining_cost_p90": 100
    }
  ],
  "correlations": [
    {
      "correlation": 0.8,
      "id": "shared-delivery-shock",
      "left_component_id": "backend",
      "right_component_id": "migration"
    }
  ],
  "maximum_budget_breach_probability": 0.5,
  "seed": 17,
  "simulation_draws": 5000
}

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.

{
  "component_tail_contributions": [
    {
      "actual_cost_to_date": 10,
      "component_id": "backend",
      "eac_tail_contribution": 137.7581,
      "simulated_expected_remaining_cost": 57.8136,
      "submitted_remaining_cost_p50": 50,
      "submitted_remaining_cost_p90": 100
    },
    {
      "actual_cost_to_date": 10,
      "component_id": "migration",
      "eac_tail_contribution": 137.2535,
      "simulated_expected_remaining_cost": 57.7918,
      "submitted_remaining_cost_p50": 50,
      "submitted_remaining_cost_p90": 100
    }
  ],
  "decision": "estimate_at_completion_clears_governed_risk_gates",
  "failed_gates": [],
  "guardrails": [
    "P50/P90 inputs must be calibrated on immutable estimates and later resolved component costs from this organization; Git activity is not a cost distribution.",
    "The Gaussian copula preserves the submitted pair dependence but one lognormal component cannot represent multimodal scope change, cancellation, or omitted discrete risk events.",
    "Correlation uplift is a model comparison, while component tail contribution is risk accounting; neither is causal credit, blame, or an employment signal.",
    "This bottom-up forecast complements rather than silently replaces earned-value CPI/SPI, contractual commitments, or finance-approved rebaselines."
  ],
  "method": "bottom_up_gaussian_copula_lognormal_eac_simulation_v1",
  "simulation_diagnostics": {
    "antithetic_draws": true,
    "dkw_cdf_probability_error_bound_95": 0.0192,
    "minimum_correlation_eigenvalue": 0.2,
    "seed": 17,
    "submitted_p50_is_lognormal_median": true,
    "submitted_p90_sets_lognormal_scale": true
  },
  "summary": {
    "actual_cost_to_date": 20,
    "budget_at_completion": 200,
    "component_count": 2,
    "correlation_uplift_at_p90": 16.0894,
    "cvar_eac": 275.0116,
    "expected_eac": 135.6055,
    "p10_eac": 72.9268,

How it works

Forecasting & survival — Estimate when something completes or fails, with censoring and unresolved work handled honestly rather than dropped.

1. 1 Freeze the scope baseline and actual cost to date, calibrate each component remaining-cost median and p90 on comparable resolved work, and assemble a point-in-time positive-semidefinite dependence matrix from shared delivery and market shocks.
2. 2 Convert each p50/p90 pair into a lognormal marginal, join marginals with the Gaussian copula, use seeded antithetic Monte Carlo to simulate total EAC, and compare its p90 with the same marginals under independence.
3. 3 Report EAC quantiles, breach probability, CVaR, a DKW finite-draw probability bound, and component tail accounting; apply probability and CVaR gates without overwriting earned-value, contractual, or rebaseline evidence.

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": "turn bottomup component actuals and locally" }
  → finds "estimate_estimate_at_completion_distribution"

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

gitrevio_capability_run
  { "capability_id": "estimate_estimate_at_completion_distribution", "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 Delivery forecasting & commitments →