Calibrate business case forecasts

What it's for

Learns whether this organization systematically over- or under-forecasts business value, while shrinking sparse categories and proving the correction on later resolved cases before touching live cases.

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.

{
  "current_cases": [
    {
      "category": "growth",
      "forecast_value": 200,
      "id": "current-growth"
    },
    {
      "category": "efficiency",
      "forecast_value": 150,
      "id": "current-efficiency"
    }
  ],
  "interval_probability": 0.8,
  "maximum_coverage_error": 0.2,
  "minimum_category_cases": 5,
  "resolved_cases": [
    {
      "category": "growth",
      "forecast_value": 100,
      "id": "resolved-case-0",
      "realized_value": 140,
      "resolved_order": 0
    },
    {
      "category": "efficiency",
      "forecast_value": 101,
      "id": "resolved-case-1",
      "realized_value": 80.80000000000001,
      "resolved_order": 1
    },
    {
      "category": "growth",
      "forecast_value": 102,
      "id": "resolved-case-2",
      "realized_value": 142.79999999999998,
      "resolved_order": 2
    },
    {
      "category": "efficiency",
      "forecast_value": 103,
      "id": "resolved-case-3",
      "realized_value": 82.4,
      "resolved_order": 3

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": [
    "Forecast and realized values are strictly positive, incremental, comparable, and expressed in one governed value definition, currency basis, horizon, and vintage; resolved_order preserves what would have been known at deployment time.",
    "Log forecast error is sufficiently stable for a global normal residual with empirical-Bayes category location effects; sparse and unseen categories fall back toward the global correction rather than borrowing another tenant's multiplier.",
    "A calibration is supported only when it improves a proper out-of-time log score, does not worsen absolute log error, and attains the governed interval coverage tolerance; current-case outputs remain provisional when the gate fails.",
    "Calibration repairs systematic forecast distributions, not causal business cases, benefit definitions, omitted costs, strategic incentives, manipulation, or individual performance judgments."
  ],
  "category_diagnostics": [
    {
      "case_count": 30,
      "category": "efficiency",
      "category_specific": true,
      "effective_case_count": 30,
      "raw_calibration_multiplier": 0.8,
      "shrinkage_weight_on_category": 0.9667,
      "shrunk_calibration_multiplier": 0.8075
    },
    {
      "case_count": 30,
      "category": "growth",
      "category_specific": true,
      "effective_case_count": 30,
      "raw_calibration_multiplier": 1.4,
      "shrinkage_weight_on_category": 0.9667,
      "shrunk_calibration_multiplier": 1.387
    }
  ],
  "configuration": {
    "chronological_split": true,
    "interval_probability": 0.8,
    "maximum_coverage_error": 0.2,
    "minimum_category_cases": 5,
    "minimum_log_score_improvement": 0,
    "positive_value_logratio_model": true,
    "validation_fraction": 0.25
  },
  "current_case_diagnostics": [
    {
      "calibrated_mean_value": 121.2826,
      "calibrated_median_value": 121.1245,
      "calibrated_quantiles": {
        "p10": 113.4486,
        "p50": 121.1245,
        "p90": 129.3197

How it works

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

1. 1 Freeze resolved business cases with immutable forecast vintage, comparable positive forecast and realized value, category, weight, and chronological resolution order; reserve the latest cases before fitting.
2. 2 Model log(realized/forecast) with a global correction and empirical-Bayes category effects that shrink sparse categories toward the global distribution, then score the frozen future holdout against an uncalibrated baseline.
3. 3 Publish corrected median, mean, and p10/p50/p90 only with an explicit support flag, requiring proper log-score improvement, non-worse absolute log error, and governed interval coverage; unseen categories use the global fallback.

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": "calibrate positive businesscase forecasts with chronological" }
  → finds "calibrate_business_case_forecasts"

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

gitrevio_capability_run
  { "capability_id": "calibrate_business_case_forecasts", "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 Investment & portfolio choice →