Forecast onboarding time to autonomy economics

What it's for

Forecasts how long new joiners still need to reach autonomy, including the ones who haven't got there yet, and prices the delay.

Turns onboarding milestones into a board- and team-ready answer: when will this cohort likely reach locally defined autonomy, what capacity remains unavailable, and how costly is the tail?

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_episodes": [
    {
      "age_days": 10,
      "capacity_value_per_day": 100,
      "evidence_verified": true,
      "fixed_delay_loss": 500,
      "id": "current-cohort-a",
      "interruption_rate": 1,
      "mentor_hours_first_30d": 8,
      "remote_fraction": 0.25,
      "role_id": "backend",
      "task_complexity": 3
    }
  ],
  "economic_scenarios": [
    {
      "duration_multiplier": 1,
      "id": "normal",
      "probability": 0.8,
      "value_multiplier": 1
    },
    {
      "duration_multiplier": 1.25,
      "id": "adverse",
      "probability": 0.2,
      "value_multiplier": 1.5
    }
  ],
  "historical_episodes": [
    {
      "autonomy_observed": false,
      "cohort_period": 0,
      "duration_or_censor_days": 23,
      "evidence_verified": true,
      "id": "autonomy-0-0",
      "interruption_rate": 0,
      "mentor_hours_first_30d": 0,
      "remote_fraction": 0,
      "role_id": "backend",
      "task_complexity": 1
    },
    {
      "autonomy_observed": true,

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.

{
  "decision": "forecast_available",
  "episode_forecasts": [
    {
      "delay_cost_cvar": 7185.2382,
      "delay_cost_var": 6467.4229,
      "episode_ref": "current-cohort-a",
      "expected_delay_cost": 3674.9561,
      "expected_remaining_days": 28.5042,
      "observed_age_days": 10,
      "remaining_days_p10": 20.487,
      "remaining_days_p50": 27.9766,
      "remaining_days_p90": 37.7176,
      "role_fallback_used": false,
      "role_id": "backend"
    }
  ],
  "interpretation": "The model conditions remaining time on survival beyond observed age and treats mentor hours, remote fraction, interruptions, and complexity as predictive associations. It does not estimate mentor causality, rank people, or support employment decisions; unsupported forecasts remain diagnostic only.",
  "method": "right_censored_partially_pooled_lognormal_autonomy_aft_v1",
  "model_diagnostics": {
    "log_duration_sigma": 0.15,
    "random_seed": 17,
    "represented_role_ids": [
      "backend",
      "frontend"
    ],
    "ridge_penalty": 2,
    "simulation_count": 500
  },
  "portfolio_summary": {
    "current_episode_count": 1,
    "delay_cost_cvar": 7185.2382,
    "delay_cost_var": 6467.4229,
    "expected_total_delay_cost": 3674.9561,
    "expected_total_remaining_days": 28.5042,
    "tail_probability": 0.95
  },
  "truncated_episode_count": 0,
  "validation": {
    "gates": {
      "current_role_support": true,
      "duration_mae_challenge": true,
      "evidence_verified": true,
      "holdout_support": true,

How it works

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

1. 1 Reserve the latest complete cohort period; fit a right-censored log-normal AFT model with task complexity, remote share, early mentor hours, interruption rate and ridge-shrunk role effects on earlier company data.
2. 2 Challenge it against a global censored log-normal baseline on later-cohort survival log score and observed-duration MAE, require interval coverage, evidence support and represented current roles, and retain unsupported results as diagnostic only.
3. 3 Condition each current episode on having remained non-autonomous through its observed age, replay coherent duration/value scenarios with common draws, and aggregate remaining-time and delay-cost VaR/CVaR.

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 remaining timetoautonomy and delaycost tails" }
  → finds "forecast_onboarding_time_to_autonomy_economics"

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

gitrevio_capability_run
  { "capability_id": "forecast_onboarding_time_to_autonomy_economics", "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 →