Forecast governed attrition competing risks

What it's for

Replaces the site's fixed weighted attrition score and assumed exponential departure window with locally tested competing-risk probabilities and honest uncertainty that improve as real outcomes mature.

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.

{
  "as_of_ms": 1000,
  "current_subjects": [
    {
      "data_cutoff_ms": 900,
      "evidence_verified": true,
      "features": {
        "signal_a": 2,
        "signal_b": 0
      },
      "id": "candidate-high",
      "role_id": "engineer",
      "scope_id": "company",
      "supportive_use_authorized": true
    },
    {
      "data_cutoff_ms": 900,
      "evidence_verified": true,
      "features": {
        "signal_a": 0,
        "signal_b": -2
      },
      "id": "candidate-low",
      "role_id": "manager",
      "scope_id": "company",
      "supportive_use_authorized": true
    }
  ],
  "feature_names": [
    "signal_a",
    "signal_b"
  ],
  "historical_periods": [
    {
      "at_risk": true,
      "event_type_id": "voluntary_exit",
      "evidence_verified": true,
      "features": {
        "signal_a": 2,
        "signal_b": 0
      },
      "id": "period-0",
      "outcome_mature": true,
      "period_index": 0,

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": 1000,
    "feature_names": [
      "signal_a",
      "signal_b"
    ],
    "holdout_fraction": 0.2,
    "horizon_periods": 4,
    "peer_prior_strength": 0.05,
    "posterior_draws": 100,
    "random_seed": 7,
    "target_scope_id": "company",
    "validation_calendar_cutoff_ms": 81
  },
  "decision": "supportive_forecast_ready",
  "failed_gates": [],
  "forecasts": [
    {
      "any_transition_probability": {
        "median": 0.9999,
        "p05": 0.9954,
        "p95": 1
      },
      "internal_transfer_probability": {
        "median": 0.001,
        "p05": 0.0001,
        "p95": 0.0095
      },
      "involuntary_exit_probability": {
        "median": 0.0628,
        "p05": 0.0099,
        "p95": 0.2961
      },
      "median_transition_period": 1,
      "subject_ref": "candidate-high",
      "voluntary_exit_probability": {
        "median": 0.9351,
        "p05": 0.7027,
        "p95": 0.9895
      }
    },
    {
      "any_transition_probability": {

How it works

Sequential Bayesian & bandits — Learn while deciding — update beliefs as evidence arrives and choose where the next unit of effort is worth spending.

1. 1 Validate complete consecutive at-risk person-period histories and hold out the latest whole local employment spells so no subject crosses training and validation.
2. 2 Fit a regularized multinomial hazard model with training-only scaling and role effects, shrinking local coefficients toward a peer-only model without counting local evidence twice.
3. 3 Require later local log-loss improvement and voluntary-exit calibration, refit on all mature history, then propagate a diagonal Laplace posterior through cumulative competing incidence over the requested horizon.

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": "forecast voluntary departure internal transfer and" }
  → finds "forecast_governed_attrition_competing_risks"

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

gitrevio_capability_run
  { "capability_id": "forecast_governed_attrition_competing_risks", "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 People, retention & knowledge →