Underwriting · Goable

The why

Traditional vs parametric

Outdoor-business interruption is a classic claims-handling nightmare: subjective loss assessment, photo evidence, multi-month settlements. Parametric flips the contract: pre-agreed trigger, objective measurement, automatic payout. The leg that was missing was the trigger model. We ship it.

Traditional indemnity

Adjuster visits the spot
Photos + customer statements
60–90 day settlement
Loss ratio swings ±30% per season

Parametric trigger

Wind < 12kt for 4h = trigger met
ERA5 reanalysis confirms within 7 days
Same-day settlement on confirmation
Loss ratio bounded by trigger frequency

Worked example

One quote · Tarifa kite-school · August 2026

A kitesurfing school in Tarifa wants weather-cancellation cover for their August teaching season — 31 days, 4-hour windows of ≥12kt sustained wind. The premium is computed from the engine's historical exceedance + per-cell forecast skill + operator loss loading.

Quote inputs

POST /v1/underwriting/quote
{
 spot: "kitesurf-tarifa-balneario",
 trigger: {
 metric: "wind_kt",
 operator: "lt",
 threshold: 12,
 duration_hours: 4
 },
 coverage: {
 from: "2026-08-01",
 to: "2026-08-31",
 loss_given_trigger_eur: 11400
 }
}

Premium derivation€2,847.20

Forecast skill (Brier)
per-cell historical, refit nightly
0.92
Historical exceedance
ERA5 multi-decade base rate
8.4%
Tier risk multiplier
sub-spot weighted hierarchical fit
1.18
Loss given trigger
operator-supplied LGT
€11,400
Loading + IPT
actuary discretion + insurance premium tax
1.42×

Reading the premium · Base hazard cost = exceedance × LGT = 0.084 × €11,400 = €957/day-of-coverage on a perfect skill basis. Adjusted by 1/0.92 forecast-skill loading → €1,040. Tier multiplier 1.18 reflects the sub-spot's historical busy-season cancellation profile → €1,228. Loadings × 31 days, IPT included → €2,847.20 final premium. Every multiplier traces to a published metric — your actuary defends it.

How it works

Quote · bind · evaluate · settle

Three endpoints + a continuous surveillance layer. Quotes are stateless; bindings snapshot the calibration; evaluations consult the operator outcome or the ERA5 reanalysis. The drift monitor watches the engine cohort for regime shift that would invalidate future bindings.

Quote

POST /v1/underwriting/quote → multi-currency premium + confidence interval. Forecast skill × historical exceedance × loss-given-trigger × loadings, all sourced from the same calibrated curves /v1/score uses.

Bind

POST /v1/underwriting/bind locks the policy_terms snapshot — the cell, the trigger threshold, the evaluation rule, the calibration hash. Tamper-evident: any drift in the engine cohort invalidates this binding, not the next one.

Evaluate

Window expires → cron runs /v1/underwriting/evaluate → observed outcome from operator (or ERA5 reanalysis) decides settlement. Webhook fires; receipt hash logged.

Drift surveillance

The SPC charter (Statistical Process Control) tracks per-cell skill via CUSUM. Watch / warning / critical drift fires alerts before settlement breaks. NAIC 715, Solvency II Pillar 2, Lloyd's MS13 compliant.

Audit trail

Every quote and binding carries the SHA-256 cohort hash. Reproducible from the open catalog + the ERA5 source — your actuary can re-derive the loading without our cooperation.

Compliance

Built for the binding-authority application

Lloyd's MS13

Performance & oversight — drift charter doubles as the model-monitoring artefact for the binding-authority application.

Solvency II Pillar 2

ORSA-ready: the SPC charter + cohort hashes + per-cell skill estimates feed the model-governance section directly.

NAIC 715

Insurance Data Security — pseudonymous + tenant-scoped, audit log retention configurable per DPA. EU residency by default.

Reproducibility

Every quote is a function of (cohort hash, calibration version, profile maturity). Same inputs → same number, forever.

Need the SPC drift charter as a JSON-LD or PDF artefact?See the charter

Operational

Settlement is a webhook

Bind a policy → window expires → the evaluator cron fires. The settlement webhook carries the binding's policy id, the observed outcome source, the amount, and the receipt hash. Your treasury reconciles by receipt hash, not by spreadsheet.

HMAC-SHA256 · same delivery contract as score webhooks

Reasonable retry posture (5 attempts, exponential backoff)
Receipt hash binds settlement to the original binding's cohort
Idempotent on policyId (replay-safe)

{
  "event": "underwriting.policy.evaluated",
  "policyId": "pol_a7b2...",
  "data": {
    "tenantId": "tn_...",
    "binding": { "cohortHash": "8f2c…", "trigger": "wind_lt_12kt for 4h" },
    "observed": { "outcome": "trigger_met", "source": "operator_report" },
    "settlement": {
      "amountEur": 11400.00,
      "currency": "EUR",
      "paidAt": "2026-08-14T11:02:14Z"
    },
    "receiptHash": "sha256:c9a1…"
  }
}

Get in touch

Underwriting is contracted — Scale tier

The underwriting endpoints aren't self-serve. We onboard each partner with a discovery call (your portfolio, your trigger ergonomics, your reinsurance), then bind the API access to a contracted tenant on the Scale plan. Typical lead-time from first call to first quote: 3–5 weeks.

Talk to the underwriting desk Partner programme SPC drift charter