Forecast customer facing service interruption loss
Forecast customer-facing outage frequency, duration, SLA credits, interrupted revenue, churn exposure and total financial VaR/CVaR using local zero-inclusive service history, compound log-normal severity and coherent shared-dependency events.
What it's for
Turns service reliability into board-ready customer and cash exposure while preserving contract economics, shared dependency concentration and severe tails.
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.
| Field | Type | Role | Required |
|---|---|---|---|
| current_service_exposures | array of objects (17 fields) | Evidence | Yes |
| historical_service_classes | array of objects (12 fields) | Evidence | Yes |
| max_detail_rows | integer ≥ 1, ≤ 500 | Numerical control | Optional |
| minimum_history_periods | integer ≥ 1, ≤ 10000 | Your calibration | Optional |
| minimum_loss_observations | integer ≥ 1, ≤ 100000 | Your calibration | Optional |
| minimum_recovery_observations | integer ≥ 1, ≤ 100000 | Your calibration | Optional |
| random_seed | integer ≥ 0, ≤ 4294967295 | Your calibration | Optional |
| scenarios | array of objects (7 fields) | Evidence | Yes |
| shared_dependency_groups | array of objects (6 fields) | Evidence | Yes |
| simulation_count | integer ≥ 1000, ≤ 200000 | Your calibration | Optional |
| tail_probability | number > 0, < 1 | Your calibration | Optional |
Each current_service_exposures
record
| Field | Type | Required |
|---|---|---|
| customer_churn_loss_per_outage | number (≥ 0) | Yes |
| evidence_verified | boolean | Yes |
| forecast_service_hours | number (≥ 0) | Yes |
| id | string (non-empty) | Yes |
| loss_prior_log_mean | number | Yes |
| loss_prior_log_sd | number (> 0, ≤ 5) | Yes |
| loss_prior_strength | number (> 0) | Yes |
| outage_rate_prior_rate | number (> 0) | Yes |
| outage_rate_prior_shape | number (> 0) | Yes |
| preventive_control_effectiveness | number (≥ 0, ≤ 1) | Yes |
| recovery_prior_log_mean_minutes | number | Yes |
| recovery_prior_log_sd | number (> 0, ≤ 5) | Yes |
| recovery_prior_strength | number (> 0) | Yes |
| revenue_loss_per_outage_minute | number (≥ 0) | Yes |
| service_class | string (non-empty) | Yes |
| shared_dependency_group_id | string (non-empty) | Yes |
| sla_credit_per_outage_minute | number (≥ 0) | Yes |
{
"current_service_exposures": [
{
"customer_churn_loss_per_outage": 20000,
"evidence_verified": true,
"forecast_service_hours": 720,
"id": "payments",
"loss_prior_log_mean": 10.819778284410283,
"loss_prior_log_sd": 0.5,
"loss_prior_strength": 5,
"outage_rate_prior_rate": 720,
"outage_rate_prior_shape": 1,
"preventive_control_effectiveness": 0.3,
"recovery_prior_log_mean_minutes": 3.8066624897703196,
"recovery_prior_log_sd": 0.4,
"recovery_prior_strength": 5,
"revenue_loss_per_outage_minute": 500,
"service_class": "tier-one",
"shared_dependency_group_id": "cloud-region",
"sla_credit_per_outage_minute": 100
}
],
"historical_service_classes": [
{
"evidence_verified": true,
"exposure_service_hours": 720,
"financial_loss_log_squared_sum": 117.06760212379633,
"financial_loss_log_sum": 10.819778284410283,
"financial_loss_observation_count": 1,
"id": "tier-one-history-0",
"period": 0,
"recovery_log_minutes_squared_sum": 14.490679311024369,
"recovery_log_minutes_sum": 3.8066624897703196,
"recovery_observation_count": 1,
"service_class": "tier-one",
"verified_outage_count": 1
}
],
"minimum_history_periods": 1,
"minimum_loss_observations": 1,
"minimum_recovery_observations": 1,
"random_seed": 43,
"scenarios": [
{Truncated for display — the full payload is 80 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.
{
"assumptions": [
"History is consecutive and zero-inclusive; verified outage, recovery and finance-loss maturity remain comparable inside each aggregate service class, and only information known at forecast origin enters the cohort.",
"One coherent scenario drives frequency, duration, loss and common dependency events; customer credits, revenue interruption, churn exposure and unique shared-group loss are finance/contract-owner inputs and common loss is counted once."
],
"counts": {
"current_service_exposures": 1,
"history_periods": 1,
"scenarios": 2,
"service_classes": 1,
"shared_dependency_groups": 1,
"simulations": 1000
},
"decision": "review_customer_facing_interruption_and_cash_tail",
"limitations": [
"This forecasts aggregate represented service/customer exposure, not a service-level promise, customer-specific legal conclusion, vendor fault, incident cause or guaranteed recovery time.",
"Unreported incidents, changing detection, contract enforceability, nonstationary architecture, insurance/recovery and losses outside the submitted perimeter can make the tail incomplete."
],
"method": "gamma_poisson_compound_lognormal_common_dependency_customer_loss_v1",
"portfolio_forecast": {
"expected_customer_contract_loss": 60957.8814,
"expected_outage_minutes": 58.1031,
"expected_outages": 0.82,
"expected_shared_dependency_loss": 13132.6155,
"expected_total_financial_loss": 122080.4334,
"financial_loss_conditional_value_at_risk": 795250.925,
"financial_loss_value_at_risk": 545142.5295,
"probability_any_outage": 0.49,
"tail_probability": 0.05
},
"reproducibility": {
"random_seed": 43,
"scenario_ids": [
"base",
"stress"
]
},
"service_forecasts": [
{
"expected_customer_contract_loss": 47825.2659,
"expected_direct_financial_loss": 61122.552,
"expected_outage_minutes": 52.3754,
"expected_outages": 0.82,
"history_periods": 1,Truncated for display — the full payload is 53 lines.
How it works
Forecasting & survival — Estimate when something completes or fails, with censoring and unresolved work handled honestly rather than dropped.
- 1 Fit Gamma-Poisson verified-outage rates by stable aggregate service class from consecutive zero-inclusive service-hour histories.
- 2 Pool local positive recovery-duration and finance-loss log moments, simulate compound individual incidents, and translate outage minutes through contract-owner SLA credit, revenue interruption and churn inputs.
- 3 Share one scenario across frequency, duration and loss, simulate one common event per dependency group, count unique shared loss once, and report expected loss, VaR/CVaR and support abstention.
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
- Training examples precede their outcomes, censoring and unresolved work are represented, and deployment populations remain comparable to validation cohorts.
- Incident detection, recovery maturity, financial perimeter, service classes and forecast exposure remain comparable; contract/customer values are enforceable owner-supplied inputs rather than inferred from engineering activity.
- A predictive interval or risk estimate is not a deadline promise, causal explanation, or individual-performance judgment.
- The forecast is aggregate scenario-conditioned exposure—not an SLA promise, customer-specific legal advice, vendor blame, incident-cause inference or guaranteed recovery time.
Minimum evidence
- historical_service_classes: required and organization-defined
- current_service_exposures: required and organization-defined
- shared_dependency_groups: required and organization-defined
- scenarios: required and organization-defined
How to validate it
Use chronological train/calibration/test windows, compare proper scores and decision value with a simple baseline, and recalibrate only from outcomes resolved after prediction time.
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
- calendar/class service-hour spine left-joined to complete verified incidents, mature recovery and finance outcomes, then joined to current services, canonical shared dependencies and effective customer-contract economics
- service-class/incident/recovery/maturity rules, exposure clock, priors, forecast horizon, SLA-credit enforceability, revenue/churn perimeter, shared-loss uniqueness, preventive effect, scenarios and tail appetite
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 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 customerfacing outage frequency duration sla" }
→ finds "forecast_customer_facing_service_interruption_loss"
gitrevio_capability_describe
{ "capability_id": "forecast_customer_facing_service_interruption_loss" }
→ returns the input schema and agent guidance shown on this page
gitrevio_capability_run
{ "capability_id": "forecast_customer_facing_service_interruption_loss", "arguments": { ... } }
→ returns the result shown aboveWorks in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.
Related tools
Audit service continuity recovery evidence integrity
Audit whether each critical service has a current, independently reviewed recovery plan whose complete capability/dependency path, backup, restore, failover, communications, RTO and RPO were proven in a recent production-representative exercise.
Optimize service continuity investment portfolio
Choose one production-exercised continuity posture per service-risk unit by maximizing retained business value minus direct/common interruption loss, full cost and CVaR under RTO, RPO, residual-risk, control, dependency, budget and resource constraints.
Aggregate risk register copula
Aggregate risk-register occurrence and lognormal severity marginals through a validated Gaussian copula into expected loss, VaR, CVaR, dependence amplification, and tail shares.
Allocate restless bandit interventions
Allocate scarce recurring interventions across evolving Markov units with Whittle indices, explicit indexability checks, and paired policy simulation.
Attribute commercial dependency tail loss
Calculate expected loss, VaR and CVaR for commercial value concentrated in shared technical components, then allocate every modeled tail-loss dollar exactly once across components with normalized negative-log survival hazard rather than overlapping leave-one-out sensitivities.
Audit commercial resilience claim integrity
Audit resilience ROI claims against a unique commercial-source to technical-component graph: recompute each action's avoided loss under joint failure scenarios, cap support at graph-derived value, detect duplicate effects, probability drift and weak evidence, and prevent overlapping component benefits from being sold twice.