Build a Windstorm Damage Claim Ledger and Evidence Reconciler

Reconcile commercial windstorm damage repair receipts against NOAA hourly station gust observations. Formulate structured settlement audit dossiers and explanatory claimant notices.

Claim Preparation Specialist

Ingest claimant repair ledgers containing roof repair invoices, tarpaulin purchases, and reported incident dates.

Retrieve observed wind gust velocities from nearest NOAA meteorological stations via Access Data Service.

Correlate recorded gust peaks against the threshold required for asserted structural component failure.

Draft an itemized ledger reconciliation summary and claimant evidence request letter.

Dual-pane review interface highlighting claimed repair costs against hourly gust velocity charts.

Query NOAA Access Data Service v1 API with stations, startDate, endDate, and dataTypes=WSFG, backed by test JSON fixtures. Starter examples below are synthetic. They are not downloaded records.

NOAA climate services API Uses the NOAA Access Data Service API endpoint parameters to query surface weather data types for specified stations and dates.

dspy, requests, pandas. Optimizes prompt signatures to ensure rigorous, evidence-grounded ledger reconciliation and standardized claimant response letters. Local or hosted LLM with strong few-shot constraint adherence and markdown ledger formatting.

Submit runnable source, pinned dependencies, setup commands, a working interface, example outputs, and a recorded end-to-end demo. Include tests and an action log.

Matches reported repair dates to NOAA station wind gust records exceeding structural damage thresholds.

Segregates expenses incurred outside verified storm wind windows into pending-evidence categories.

Drafts transparent claimant letters requesting missing itemized proof without issuing final policy denials.

```json [ { "input": { "ledger_id": "LDG-WIND-104", "loss_date": "2024-06-12", "station_id": "USW00012839", "max_gust_recorded_knots": 58, "claimed_repairs": [ { "line_id": 1, "description": "membrane_roofing_patch", "amount_usd": 8500, "date": "2024-06-13" }, { "line_id": 2, "description": "interior_repaint", "amount_usd": 3200, "date": "2024-06-25" } ] }, "expected_output": { "ledger_id": "LDG-WIND-104", "wind_evidence_verified": true, "supported_lines": [ 1 ], "flagged_lines": [ 2 ], "claimant_letter": "Peak gusts of 58 knots verified on 2024-06-12. Roof repairs aligned. Interior repaint on 2024-06-25 flagged for secondary water damage correlation." } }, { "input": { "ledger_id": "LDG-WIND-105", "loss_date": "2024-05-02", "station_id": "USW00012839", "max_gust_recorded_knots": 18, "claimed_repairs": [ { "line_id": 1, "description": "siding_replacement", "amount_usd": 6400, "date": "2024-05-03" } ] }, "expected_output": { "ledger_id": "LDG-WIND-105", "wind_evidence_verified": false, "supported_lines": [], "flagged_lines": [ 1 ], "claimant_letter": "Station recorded maximum gust of 18 knots, below wind damage threshold. Please submit localized anemometer or engineering reports." } } ] ```

Performs factual ledger reconciliation; does not execute insurance claims or legally dispute coverage. Use mock external writes. Require human review before any real deployment.