docs / api-reference

API Reference

Stripe-grade endpoint documentation. Every parameter is statically typed, every error code is documented, and every snippet is copy-pasteable.

POST/api/public/v1/land-intel/assess

Assess a coordinate

The primary scoring endpoint. Computes the 0–100 Rootfifteen score for a single GPS point. Free, public, no API key required. Accepts both GET (query params) and POST (JSON body).

BODY PARAMETERS

latitudeFloat [-90.0, 90.0]required

WGS84 decimal degrees. Rounded to 6 fractional digits.

longitudeFloat [-180.0, 180.0]required

WGS84 decimal degrees.

modeEnum<"agriculture"|"infrastructure">optional

Selects the engine specification used to compute the score. Defaults to "agriculture".

buffer_metersInt [0, 5000]optional

Reserved for future use. Default 0 (point sample).

ERROR STATES

400

invalid_geometry

lat/long missing or outside the WGS84 valid range.

400

invalid_json

POST body is not valid JSON.

503

upstream_unavailable

An upstream observation feed is momentarily unreachable. Retry with backoff.

# Free · public · no API key required
curl -X POST https://rootfifteen.lovable.app/api/public/v1/land-intel/assess \
  -H "Content-Type: application/json" \
  -d '{
    "latitude": 19.0760,
    "longitude": 72.8777,
    "mode": "infrastructure"
  }'

# GET form (handy for quick checks):
curl "https://rootfifteen.lovable.app/api/public/v1/land-intel/assess?latitude=19.076&longitude=72.8777&mode=infrastructure"

RESPONSE · 200 OK

response.assess.json

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
{
  "request_id": "req_3kx9aLp2qWmZb1c2",
  "status": "ok",
  "engine": "infrastructure.v2.8",
  "coordinates": { "latitude": 19.076, "longitude": 72.8777 },
  "buffer_meters": 0,
  "mode": "infrastructure",
  "rootfifteen_score": 64.18,
  "classification": "MARGIN_CAUTION",
  "components": {
    "climate": 71.4,
    "terrain": 92.0,
    "georisk": 41.0
  },
  "observations": {
    "mean_temperature_c": 27.6,
    "annual_precipitation_mm": 2143.5,
    "elevation_m": 14,
    "seismic_events_5yr": 6,
    "seismic_max_magnitude_5yr": 4.7
  },
  "generated_at": "2026-06-25T09:14:22Z",
  "latency_ms": 412
}

docs / algorithm internals

A single integer. A deep pipeline.

Rootfifteen exposes one normalized 0–100 number. Under it sits a deterministic data-fusion of temporal optical biophysics, Synthetic Aperture Radar deformation arrays, and Digital Elevation Models.

FUSION EQUATION

R₁₅ = clamp(100 · σ(w_opt · Φ(NDVI, MSAVI, EVI) + w_sar · Ψ(σ⁰, Δϕ, v̄) + w_dem · Ω(slope, aspect, NDWI)), 0, 100)

Weights are learned per mode and locked under a tamper-evident model registry hash.

Agriculture Engine

agriculture.v3.2 · Biophysical Anomaly Detector

90 – 100

OPTIMAL_CANOPY

Peak chlorophyll saturation. NDVI ≥ 0.78 with MSAVI co-saturation. Baseline for parametric crop policies.

70 – 89

HEALTHY

Vigorous canopy with no temporal anomaly. Phenology tracks the 5-year mean inside ±1σ.

50 – 69

PERTURBATION

Crop stress signature. z-score < -1.4 on NDVI stack — a 14–21 day leading indicator.

40 – 49

ELEVATED_STRESS

Persistent multi-pass anomaly. Triggers pre-claim notification on insurance webhooks.

Below 40

DEGRADED

Severe degradation or single-event catastrophe. Fires automated parametric payout.

Infrastructure Engine

infrastructure.v2.8 · Georisk & Engineering

80 – 100

STABLE

Absolute ground stability. Flat grade (<5°), zero hydro-logging, sub-millimetric deformation velocity.

50 – 79

MARGIN_CAUTION

Steep slope (>15°) or mild ground deformation (1–5 mm/yr).

30 – 49

ENGINEERING_REVIEW

Compounding signals — moderate subsidence + adverse slope. Manual geotechnical review.

Below 30

CRITICAL_HAZARD

Active subsidence (>5 mm/yr) or rapid NDWI flash-flood anomalies. Auto-flagged on civil pipelines.

docs / engine advantage

An API versus a clipboard.

How Rootfifteen's continuous pipeline compares to the legacy field-and-rig methods still embedded in insurance underwriting and civil engineering compliance.

Dimension

Rootfifteen API

Legacy Agricultural

Legacy Engineering

Acquisition

Continuous temporal ML pipeline

Subjective manual field loss-adjusters

Intermittent geotechnical bore-hole sampling

Turnaround

~80ms per HTTP request

Weeks (claim → field visit → report)

Days to weeks per bore-hole campaign

Coverage

Global · 10m × 10m grid

Per-farm site visits

Discrete point samples

Temporal density

5–6 day revisit, 5-year coherent stack

On-demand only

One-shot at survey time

Cost / observation

Sub-cent at scale

Field visits + adjuster labor

Rig mobilization + drone permits

Basis risk

Algorithmically bounded · auditable hash

High administrative basis risk

Spatial extrapolation between samples

Output

Clean 0–100 score + JSON sub-matrices

PDF report

PDF + CAD attachments