Forecast organizational change second order effects
Forecast the incremental capacity, review, knowledge, quality, backlog, recovery and financial distribution of a submitted departure, hire, restructure, reassignment, PTO, AI rollout, framework migration or contractor scenario using company-local completed episodes and a dependency DAG.
What it's for
Provides the actual calibrated Monte Carlo engine behind the What-If Simulator's departure, hiring, restructuring, reassignment, PTO, AI rollout, migration and contractor scenarios.
What you give it
Every tool is a pure function: it computes from what you pass it and reads nothing on its own. Inputs split into evidence you supply from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.
| Field | Type | Role | Required |
|---|---|---|---|
| as_of_ms | number ≥ 0 | Your calibration | Yes |
| baseline_units | array of objects (13 fields) | Evidence | Yes |
| current_change_impacts | array of objects (9 fields) | Evidence | Yes |
| dependency_edges | array of objects (6 fields) ≥ 0 items | Evidence | Yes |
| historical_change_periods | array of objects (11 fields) | Evidence | Yes |
| horizon_periods | integer ≥ 2, ≤ 520 | Your calibration | Yes |
| max_detail_rows | integer ≥ 1, ≤ 500 | Numerical control | Optional |
| minimum_local_episodes | integer ≥ 2, ≤ 10000 | Your calibration | Optional |
| prior_strength | number ≥ 0 | Your calibration | Optional |
| random_seed | integer ≥ 0, ≤ 4294967295 | Your calibration | Optional |
| recovery_service_fraction | number > 0, ≤ 1 | Your calibration | Optional |
| scenarios | array of objects (8 fields) | Evidence | Yes |
| simulation_count | integer ≥ 1000, ≤ 200000 | Your calibration | Optional |
Each baseline_units
record
| Field | Type | Required |
|---|---|---|
| baseline_capacity_per_period | number (≥ 0) | Yes |
| baseline_cost_per_period | number (≥ 0) | Yes |
| baseline_defect_probability | number (≥ 0, ≤ 1) | Yes |
| baseline_demand_per_period | number (≥ 0) | Yes |
| baseline_knowledge_coverage | number (≥ 0, ≤ 1) | Yes |
| baseline_review_capacity_per_period | number (≥ 0) | Yes |
| delay_cost_per_backlog_item | number (≥ 0) | Yes |
| evidence_verified | boolean | Yes |
| id | string (non-empty) | Yes |
| loss_per_defect | number (≥ 0) | Yes |
| opening_backlog | number (≥ 0) | Yes |
| reviews_per_output | number (> 0) | Yes |
| value_per_completed_output | number (≥ 0) | Yes |
{
"as_of_ms": 17280000000,
"baseline_units": [
{
"baseline_capacity_per_period": 12,
"baseline_cost_per_period": 4000,
"baseline_defect_probability": 0.03,
"baseline_demand_per_period": 10,
"baseline_knowledge_coverage": 0.9,
"baseline_review_capacity_per_period": 12,
"delay_cost_per_backlog_item": 200,
"evidence_verified": true,
"id": "platform",
"loss_per_defect": 2000,
"opening_backlog": 0,
"reviews_per_output": 1,
"value_per_completed_output": 1000
},
{
"baseline_capacity_per_period": 10,
"baseline_cost_per_period": 5000,
"baseline_defect_probability": 0.04,
"baseline_demand_per_period": 9,
"baseline_knowledge_coverage": 0.85,
"baseline_review_capacity_per_period": 10,
"delay_cost_per_backlog_item": 300,
"evidence_verified": true,
"id": "checkout",
"loss_per_defect": 3000,
"opening_backlog": 0,
"reviews_per_output": 1,
"value_per_completed_output": 1500
}
],
"current_change_impacts": [
{
"active_period_count": 3,
"change_class_id": "senior-owner-departure",
"change_type_id": "departure",
"effect_scale": 1,
"evidence_verified": true,
"id": "platform-departure",
"incremental_cost_per_period": 500,
"start_period": 0,Truncated for display — the full payload is 280 lines.
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.
{
"configuration": {
"as_of_ms": 17280000000,
"horizon_periods": 6,
"minimum_local_episodes": 5,
"prior_strength": 8,
"random_seed": 71,
"recovery_service_fraction": 0.95
},
"decision": "change_effect_forecast_available",
"forecast": {
"incremental_cost_p50": 1500,
"incremental_value": {
"p10": -21190.9888,
"p50": -3588.9534,
"p75": -3465.1937,
"p90": -3366.8553,
"probability_positive": 0
},
"loss_cvar_95": 22112.0698,
"loss_value_at_risk_95": 21672.1164,
"periods": [
{
"backlog_excess_p50": 3.0005,
"cumulative_incremental_value_p50": -4690.5771,
"period": 1,
"throughput_ratio_p50": 0.8421,
"throughput_ratio_p90_downside": 0.7334
},
{
"backlog_excess_p50": 3.0925,
"cumulative_incremental_value_p50": -6116.5749,
"period": 2,
"throughput_ratio_p50": 0.9934,
"throughput_ratio_p90_downside": 0.8628
},
{
"backlog_excess_p50": 0.6852,
"cumulative_incremental_value_p50": -4231.5909,
"period": 3,
"throughput_ratio_p50": 1.1235,
"throughput_ratio_p90_downside": 0.9849
},
{Truncated for display — the full payload is 122 lines.
How it works
Causal inference & experiment design — Separate what a change caused from what merely moved alongside it.
- 1 Require complete verified local episode-period histories for every submitted change class and active period; transform capacity, review, knowledge, demand and defect-odds responses to log space and shrink sparse local moments toward the organization's period-specific portfolio history.
- 2 Draw correlated response vectors and one coherent operating scenario per simulation, then replay the changed and unchanged systems under identical futures through the acyclic dependency graph, review constraints, nonlinear throughput and backlog recurrence.
- 3 Price output, defect, backlog and incremental cost on both paths; report paired incremental value, VaR/CVaR, conditional recovery, time paths, unit spillovers and calibration support, or abstain before simulation when local support is incomplete.
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.
Assumptions & guardrails
- Assignment or adoption timing, interference policy, overlap, attrition, and outcome availability match the estimand encoded by the method.
- Completed episodes are comparable to the current change class, ratios are measured consistently, change-response dynamics are stable over the horizon, the dependency DAG and lag fractions are adequate, and scenarios form one coherent joint law.
- A causal label is warranted only when design and diagnostic requirements pass; otherwise treat the result as descriptive or abstain.
- This is a conditional paired operational forecast, not a causal or personal-performance claim, and it does not authorize employment, monitoring, procurement, rollout or migration decisions.
Minimum evidence
- historical_change_periods: required and organization-defined
- baseline_units: required and organization-defined
- current_change_impacts: required and organization-defined
- dependency_edges: at least 0 rows/items
- scenarios: required and organization-defined
- horizon_periods: required and organization-defined
- as_of_ms: required and organization-defined
How to validate it
Preserve the assignment/adoption design and validate overlap, pre-period or placebo diagnostics, attrition, interference policy, and cluster-level uncertainty; never tune on the estimated effect.
Calibrating it to your org
Same for everyone
The mathematical kernel, validation rules, method version, and JSON output semantics are organization-independent; no tenant-trained coefficients or company benchmark is embedded in the function.
Specific to you
- verified episode-period response panel classified by locally governed change class, plus one frozen aggregate unit graph joining capacity, review demand, work demand, backlog, knowledge, defect, output value, delay loss, defect loss and cost at the forecast origin
- change class and episode boundary, period cadence, comparable baseline, outcome maturity, effect scale, local-history/shrinkage gates, dependency direction/fraction/lag, recovery threshold, horizon, scenario probabilities, currency and value/loss/cost basis
Calibration workflow
- 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
- 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 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
- 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 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
- 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 20 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": "forecast the incremental capacity review knowledge" }
→ finds "forecast_organizational_change_second_order_effects"
gitrevio_capability_describe
{ "capability_id": "forecast_organizational_change_second_order_effects" }
→ returns the input schema and agent guidance shown on this page
gitrevio_capability_run
{ "capability_id": "forecast_organizational_change_second_order_effects", "arguments": { ... } }
→ returns the result shown aboveWorks in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.