Closed Beta

DeepFind Hunter is in closed beta and not yet publicly available. Some capabilities are labeled Beta while their external providers are being configured. Request tester access →

DeepFind Hunter

Legal

Data Sources & Attribution

This page covers the data sources, attribution requirements, and licensing for DeepFind Hunter's LiDAR mapping features — layered terrain viewing, offline downloads, and AI LiDAR Analysis. It mirrors the in-app Data Sources & Attributions screen.

Base map (worldwide)

OpenStreetMap

The app's default Standard layer uses OpenStreetMap tiles and is the only layer with worldwide street-level coverage. OpenStreetMap data is © OpenStreetMap contributors, available under the Open Database License; the rendered tiles are served under the OpenStreetMap Foundation's tile usage policy.

Attribution shown in-app: "© OpenStreetMap contributors."

Source: openstreetmap.org/copyright

United States topographic and imagery layers

USGS National Map basemaps

Three layers — USGS Shaded Relief, USGS Imagery + Topo, and USGS Topo — are served from the USGS National Map basemap services. They are public-domain U.S. government works and they cover the United States only: outside the US these layers render no data, which is a limitation of the source rather than a restriction we impose.

Source: basemap.nationalmap.gov

United States public elevation data (primary)

USGS 3D Elevation Program (3DEP) / The National Map

Bare-earth Digital Elevation Model (DEM) tiles power the LiDAR hillshade/terrain layer, offline downloads, and the AI LiDAR Analysis candidate detection. This data is public domain — a U.S. government work with no copyright.

Attribution shown in-app: "USGS The National Map: 3D Elevation Program. USGS EROS Center: GMTED2010."

Source: usgs.gov/3d-elevation-program

3DEP is a U.S. dataset and coverage varies by area. The app performs a real coverage check rather than assuming a rendered tile means LiDAR coverage; areas without 3DEP data surface "No 3DEP LiDAR coverage here."

Fallback terrain layer

Esri World Hillshade

A global hillshade used only as a fallback where USGS 3DEP is not available. It is subject to Esri's terms of use and is not public domain. Shown in-app as "Esri, Maxar, Earthstar Geographics, and the GIS User Community."

Source: Esri World Hillshade on ArcGIS

Layers with no data source attached

Being explicit about the gaps: these future placeholders have no tile source configured, and production Android builds hide them:

  • Historical Map Overlay — Coming Soon — not included in the current Android beta. It is awaiting a licensed or public-domain historical map source.
  • Parcel Boundaries — Coming Soon — not included in the current Android beta. It is awaiting a parcel data provider; the current app cannot look up who owns a parcel.
  • Bare-earth terrain rendering — a separate catalogue entry with no source attached. We do not present it as a capability anywhere on this site or in our marketing.

AI LiDAR Analysis pipeline

AI LiDAR Analysis (a Pro, Beta feature) is a hybrid pipeline:

  • Classical computer-vision candidate detection fetches the raw 3DEP float DEM for the selected area, computes geomorphometric derivatives, and runs conservative heuristic detectors that emit candidate features (depressions, cellar holes, ditches/trenches, old roads, building foundations, terraces, mining features, former home sites, disturbed ground, historic shorelines).
  • LLM labeling sends only candidate metadata (never user identity or raw pixels) to an LLM (OpenAI) to label, classify, and explain each candidate.

Critical disclaimer: Every AI-detected feature is labeled "possible" or "suspected" — never "confirmed," "definite," or "discovered" — and carries a "requires field verification" flag. The AI output is a research aid to direct fieldwork, not a finding of fact.

OpenAI usage: only candidate metadata (geometry type, bounding box, evidence metrics, kind) is sent — no map imagery, no user PII, no location history beyond the analyzed bounding box. See OpenAI's terms of use.

Deployment status: this pipeline runs server-side, not on your device, and both halves of it — the computer-vision detection service and the labeling step — must be deployed and configured with credentials before the feature can return anything at all. That work is not complete, which is why AI LiDAR Analysis is labeled Beta. Until it is done the app reports that analysis is unavailable rather than producing a result it cannot stand behind.

Offline use

Downloaded tiles are cached namespaced by layer. USGS public-domain tiles may be cached and used offline. OpenStreetMap and Esri fallback tiles are cached under the same mechanism but remain subject to their publishers' terms, including limits on bulk downloading. Region size is capped per download for that reason, and the largest caps sit behind a higher plan than the Free and Pro tiers offered during this beta.

Privacy

LiDAR research targets, land-access records, and historical notes are stored per-user under owner-only rules. The AI analysis sends only the selected bounding box and candidate metadata to the LLM — no user identity, no location history. See our Privacy Policy.

Licensing note

All DeepFind Hunter logo and screenshot artwork is original and brand-owned. App fonts are licensed under the SIL Open Font License and the Apache License 2.0.