Every structure. Every occupant. Every lifeline. One picture before, during, and after the event. Prometheus holds a live parcel-level record of the built environment. A watch goes up, Prometheus runs the footprint: what is exposed and what will fail. Plan and prepare for all scenarios before they actually happen.
Surge and wind footprint over every structure and estimated occupant, rolled up to H3 cells, tracts and counties. Lifelines and shelters tested against modelled depth.
Hazard maps show where. Facility lists live in separate systems. Damage arrives days later on paper. Prometheus keeps structures, people, hazards and lifelines in one record and answers from it at every phase.
Every structure in the jurisdiction scored for flood depth, surge, wind, wildfire and seismic risk, with estimated occupants. Rank mitigation projects by structures and people protected per dollar. Evidence formatted for hazard mitigation plans and grant applications.
Footprints run on every advisory. Structures, occupants, shelters, hospitals, power, water and road segments tested against modelled depth and wind. A ranked list of priority cells for evacuation, sheltering and pre-positioning, hours before landfall.
Post-event imagery tasked on the worst cells at landfall + 24h. Roof and structure damage counted per parcel. Preliminary damage counts by tract and county ahead of windshield surveys.
Damaged structures ranked by severity and occupants. Repair and rebuild tracked from permits and imagery. Substantial-damage candidates flagged. Progress reported by tract, month over month.
Built and audited in the most demanding commercial use of property data. Pointed at emergency management without a new stack.
3,000+ counties of parcel coverage. Every building footprint keyed to its parcel. Occupancy and use classified. Estimated occupants per structure from structure count, occupancy class and census density.
Diablo · DerechoFlood depth and frequency, storm surge, hurricane wind, hail, wildfire with defensible space, earthquake. Structure-level, not zone-level. Grounded in HAZUS and National Risk Index data you already use.
Vigil · Pyre · RiskRankLive hazard feeds stream in through Jetstream. Deterministic footprints run at T−72h, T−48h and T−24h and re-run as the track shifts. Structures, occupants and lifelines intersected. Results pushed to the EOC.
Jetstream · VigilPost-event imagery tasked on priority cells. Roof and structure damage assessed per parcel. Where satellites cannot fly, Neural Aerospace's KittyHawk platform tasks the imagery: sensor-agnostic, GPS-denied navigation, 50 kg / 50 min / 50 km class.
Earth Observatory · Neural AerospaceHospitals, shelters, power, water, communications and road segments tested against modelled depth and wind. Nearest passable route computed for every critical facility. Rolled up by community lifeline.
Con-QGAIA turns the fifteen questions in a decision brief into queries over the record and drafts the brief with every number traced to its source. Analysts stay in charge. The brief ships in hours.
GAIARuns inside the planning cycle, the EOC and the recovery office you already operate.
Define the area and the scenario. The footprint runs over every structure and occupant estimate. Lifelines and shelters are tested. GAIA drafts the brief with sources. Re-runs on the next advisory and shows what changed.
Ask the map a questionAsk in plain language. Every answer shows its sources.
Occupants from structure count, occupancy class and tract density. Shelters from your list. Distance by road. Ranked list exported for evacuation planning.
Zone maps, facility lists and damage surveys each answer one question, in a different system, on a different timeline. One record answers all of them at every phase.
| Dimension | Current tools | With Prometheus |
|---|---|---|
| Resolution | Zones, tracts and county-level estimates. | Every structure and its estimated occupants, rolled up to any boundary. |
| Timing | Exposure estimated after landfall. | Footprint run on every advisory from T−72h. Re-run as the track shifts. |
| Damage | Windshield surveys over days. | Imagery tasked at T+24h. Damage per parcel. Counts by tract before teams deploy. |
| Lifelines | Facility lists maintained separately. | Facilities and road segments tested against modelled depth. Nearest passable route computed. |
| Mitigation | Projects justified one at a time. | Every project modelled against the same record. Ranked by people and structures protected per dollar. |
| Transparency | Models with methods locked away. | Source, date, method version and confidence on every number. Same inputs, same output. |
The platform runs today for organisations that are audited on every number. Where a figure is an internal benchmark, it says so.
H3 cells in one Category 4 decision-brief scenario over a coastal metro.
counties of parcel coverage. 99% of the U.S. population. Refreshed quarterly.
properties resolved in 36 minutes in a portfolio audit.
California structures scored by a 12-factor wildfire model.
Source, method, rule, decision-maker. Nothing on the platform produces a number it cannot explain.
Each value shows which source produced it, when it was observed, and where data is missing.
Observed, inferred and modelled values are labelled. Occupant estimates carry their method version and confidence.
Thresholds and priority rules are yours. Prometheus ranks; the incident commander decides. No automated tasking without your rule.
Dedicated or government-cloud deployment by SOW. API, batch and streaming. Data stays under your control.
Open the example record. Then bring your own scenario.
Government engagements are priced by task order or SOW. One credit meters consumption underneath, so cost is transparent to the programme office.
Transparent to the programme office. No platform fee, no module fees, no seat pricing.
You pay for footprints you run and structures that changed, not for watching.
Dedicated environment, data handling and support defined with the agency or prime.
Your scenario, your lifeline list, the real platform.
Ask for the rate card and a target-account scenario built on your own volumes.
Pick a jurisdiction and a scenario. We run the fifteen questions over the live record and show every source.