Type an address. Freshet pulls the six federal datasets that govern its flood risk, runs standard civil-engineering methods over them, and returns a sourced 1–10 screening score — with a Pre-Engineering Brief to back it up.
scroll to take the reading ↓The gauge is anchored to 6001 Memorial Dr — parcel matched, referenced to its contributing watershed, Buffalo Bayou, HUC 12040104. Ground elevation 49.7 ft. The staff reads zero.
FEMA, USGS, NOAA and the Army Corps report in parallel. The finished floor sits 4.3 ft below the base flood elevation — the first mark on the staff. Every reading keeps its source.
Each weighted dimension marks the staff — Terrain 8.8 and Floodplain 8.4 sit high in the red; Hydrologic 6.2 and Infrastructure 5.6 hold the middle. The waterline rises between them.
The level settles at 7 / 10 — over the high-water line. Tier: High · PE escalation recommended. A pre-engineering brief is issued at the mark.
Pick one of three real screenings and replay it start to finish — including the Dallas parcel that looks safe on the FEMA map and isn’t.
These six datasets are queried live for every screening, and each number keeps the endpoint and timestamp it came from — so any figure in the file can be defended later.
The same methods a licensed engineer would apply by hand, run automatically on your parcel’s data — with every intermediate result open for inspection.
The address is geocoded, matched to its parcel, and tied to the watershed that drains to it.
All six federal sources are queried at once, and every value keeps a record of where it came from.
The engineering methods run on that data and roll up into four weighted risk dimensions.
You get a single 1–10 screening score with every input sourced, plus a Pre-Engineering Brief you can generate on demand (~30 s).
Screening a whole pipeline? Upload a CSV and Freshet works through up to 100 addresses, flags the risky ones, and returns a one-page summary you can forward as-is.
The score is never a black-box number — it breaks down into four weighted dimensions and a list of findings, each one tied to the source it came from.
“The free FEMA map is often fifteen years out of date, and a real engineering study costs $30K and takes weeks. Freshet sits between the two: it tells you, in under a minute, whether that study is worth commissioning at all.”
Nothing in a screening depends on data you couldn’t go verify yourself.
Every input is a citable public dataset — no scraped or proprietary black-box data.
Each value carries its endpoint and timestamp, exportable for the file.
TLS 1.2+ and AES-256. Keys and secrets stay server-side — never exposed to the browser.
Freshet screens and documents; it does not perform or seal engineering studies. When a screening needs one, you engage a licensed engineer and start them from the sourced brief.
Freshet is invite-only during early access. Tell us what you're screening and we'll set you up — usually same-day.
We'll be in touch shortly — usually the same day.
In the meantime, the
interactive demo
shows a full reading end to end.
An address is all the live product needs. This demo replays three screenings it already ran on live federal data — pick one.
Queries at screening time: FEMA NFHL · USGS NLDI · NOAA Atlas 14 · USGS 3DEP · USACE NID · USGS NWIS — then three civil-engineering methods over the results (Log-Pearson III, Froehlich dam-breach, SCS runoff).
Replay at ~10× speed — a live first-time screening takes 30–60 seconds while the six sources respond.
The live product writes an ~11-page Preliminary Flood Hazard Assessment and Drainage Study grounded in this screening's data. This sample shows an excerpt.