StormTrust presents one intelligence graph through several task-focused surfaces.
Storm Forecast → Storm Alert → Storm Radar → Storm Map → Storm Impact → Storm Report connects preparation and warnings to inspectable evidence, a storm timeline, geographic priorities, and a reviewed field report. The first Storm Report release can precede parcel matching and Forecast ingestion. Platform and data owns these capability explanations and their shared contracts. The descriptions below summarize the broader product surfaces.
Storm History and Storm Replay support the whole journey: users find earlier evidence, reopen reports, and inspect observations through time. Their shared contracts also support product testing, with isolated execution and explicit fixture labels.
The Storm Map is the primary visual intelligence surface. It should answer:
- Where is severe weather occurring?
- What occurred, and what is forecast?
- Which events and properties matter?
- What evidence supports each event?
- Which areas should teams work?
- What changed over time?
Storm Forecast uses outlooks for preparation. Watches and warnings retain their separate Storm Alert semantics and publisher instructions.
Radar layers include MESH, MESH accumulations, Probability of Severe Hail (POSH), radar quality, and optional verification layers.
Observation layers include preliminary reports, reconciled reports, and field confirmations.
Derived layers include event and revision boundaries, severity, confidence, hail swaths, and cumulative exposure.
Property layers include property points, licensed parcels, commercial buildings, watched properties, impacts, and priorities.
Operations layers include territories, deployment zones, assignments, completed visits, and unresolved opportunities.
An event result exposes stable identity, status, timing, geography, maximum estimated hail, severity, confidence, evidence labels, current revision, revision history, sources, property impacts, alert state, and report state.
A property surface summarizes address, geometry, parcel, structure, roof attributes, permitted provider metadata, watched status, event exposures, latest impact, contact eligibility, actions, reports, photos, and customer relationships.
Users can monitor individual properties, customer portfolios, installed roofs, commercial portfolios, territories, neighborhoods, ZIP codes, and counties.
An alert represents a policy-qualified, notification-worthy condition. It records triggering evidence, threshold, monitor, recipient, event revision, time, channel, delivery state, acknowledgment state, and deduplication key.
Search should support address, city, county, state, ZIP code, coordinates, storm event, customer, property, business, territory, report, and historical date.
Common filters include event timing, hail size, confidence, evidence class, severity, geography, radius, territory, property use, roof attributes, watched status, assignment, contact, inspection, and opportunity status.
An event timeline combines source and operational activity:
Forecast issued
Warning issued
MRMS hail signal begins
Event created
Boundary revised
Ground report received
Alert delivered
Crew deployed
Property visited
Inspection completed
Campaign approved
Lead received
Outcome recordedThe timeline distinguishes external source timestamps from StormTrust processing timestamps.
Report creation begins from a known storm, property, portfolio, date range, or evidence package. For the report model, refer to Reports and evidence.