ICHOR · OS
ICHOR / Domains / Site & market readiness
D-02 · Site & market readiness Live
Cardo

Know if it can be built before you commit a dollar.

Cardo is the ICHOR engine for site and market readiness. It fires thirty authoritative government feeds on one parcel and returns an A to F readiness grade, and it is honest, on the memo, about what it has not verified.

The decision problem

A site looks buildable until it isn't. Power capacity, water and sewer will-serve, rail access, entitlement path, labor shed, soils, environmental and airspace overlays. Any one of them can end a project, and most do not surface until months into diligence, after real money and real time are committed.

The information exists. It is spread across dozens of public bodies and datasets that no one has the time to assemble parcel by parcel, on a deal timeline. So the risk that should have been a red flag on day one becomes a write-off in month six.

The problems the data already knew should show up on screen, not six months into escrow.

Watch it run

A readiness scorecard: a single parcel graded across sixteen factors, each one scored and sourced. The methodology and the underlying government and soil data are public and authoritative.

Cardo · A to F readiness scorecardrecorded capture · 2026
Site identity is withheld. The scoring methodology and its authoritative sources run at full fidelity.

The method, applied to readiness

The same five-step method behind every ICHOR system, run into the ground truth of site delivery. See the method in full →

01 · Source

On a single parcel, fire roughly thirty authoritative feeds at once: USDA-NRCS soils, FEMA flood, USGS slope, USFWS wetlands and habitat, HIFLD grid, FAA airspace, EPA and state environmental, Census labor, and the county parcel and zoning layers.

02 · Resolve

Reconcile them into one picture per parcel, each factor tagged for how it is known, and the will-serve items that cannot be auto-checked left unresolved rather than assumed.

03 · Decide

Score each of sixteen factors with its own function derived from the real physics or regulation, weight them for the intended use, and roll them to an A to F grade, re-normalized over only the factors that returned live data.

04 · Act

Run four knockout gates independently, so a fatal flaw renders the site NOT READY regardless of the grade, and surface the blockers and drivers for the go or no-go call.

05 · Prove

Every factor traces to its authoritative feed. The score is deterministic and reproducible, and the memo states plainly what it has not yet verified.

Sixteen factors, each computed for real

Not a checklist. Each factor has its own scoring function, built from the physics or the regulation behind it.

AirspacePure 14 CFR Part 77 geometry: nearest-runway distance, the Form 7460-1 notice height, and the imaginary-surface ceilings, layered with military noise and accident-potential zones.
Soils and geotechA weighted composite off USDA-NRCS survey over a parcel-scale polygon: shrink-swell, bearing index, water-table depth, and drainage, with a compound knockout for the worst pairs.
PowerNearest transmission distance banded by corridor voltage, upgraded by the closest substation and its nameplate, with a redundancy bonus for a second high-voltage corridor.
Buildable acreageParcel acres haircut by the multiplicative overlay of FEMA flood, measured slope, and wetland coverage, with floodway and sub-acre results treated as knockouts.
Plus rail and transload, highway access, labor shed, water rights, environmental, liquefaction, entitlement path, site cost, species and habitat, incentives, zoning, and access. Sixteen in all, re-weighted for five intended uses from big-box to data center.

What it will not fake

A skeptical developer's first objection to any AI is that it overclaims. This one is built to pre-empt it.

It tells you what it does not know.Power capacity, water and sewer will-serve, and clean title are shown as unresolved relationship gates, never a false pass, with a verify-before-close line on every memo.
Every factor is tagged for how it is known.Measured, a hard polygon or raster fact. Proxy, a distance or radius. Screening, an indicative survey. The tag is on the row, so a fact is never mistaken for an estimate.
The deal-killers are judged on their own.Zoning, geotechnical, floodway, and buildable acreage each render a site NOT READY independent of its score, because one fatal flaw is not outweighed by nine strengths.
A transparent rules engine, deterministic, with no language model in the score. Its credibility is that you can see every threshold, not a claimed accuracy number.

By the numbers

Breadth times speed: a full site-diligence memo on one parcel click, from the authoritative record.

16
development factors, each scored from its own authoritative feed
~30
live government data feeds, queried in one parallel pass
20+
named federal and state agencies, from USDA-NRCS to FAA to HIFLD
5
intended-use profiles, each reweighting the factors
4
independent knockout gates that can render a site NOT READY
One click
a full A to F readiness memo, in seconds

The entitlement layer, read every day

Cardo does not wait for the GIS map. It reads the corridor's governing bodies every day, harvests each new agenda and set of minutes, and extracts the rezones, annexations, and development agreements with the source excerpt. When a council approves an industrial rezone, the parcel scores by its new, approved zoning before the public map catches up.

governing bodies readthe corridor · daily
agenda platforms handledsix, plus the statutory notice feed
actions extractedrezone / annexation / GPA, cited
every recordsource-excerpt, append-only
effect on the gradeapproved zoning overrides the lagging map
Illustrative of the daily run. Specific parcels are not shown.

See it run on your pipeline.

ICHOR deploys one industry at a time, forward-deployed and tied to outcomes. Every inquiry is read personally.

Request access