Available now
Prototype website, concept demos, strategic positioning and public-listing research queue.
ATLAS methodology / In development
The score should explain the decision, not hide the uncertainty.
ATLAS is being designed as an evidence and workflow system for physical AI infrastructure. The objective is not to produce impressive scores from weak inputs. The objective is to make every claim, unknown and next decision explicit.
Why this matters
Infrastructure decisions fail when evidence is flattened into optimism.
Status note. This methodology describes the operating standard ATLAS is being built toward as of July 12, 2026. It is not a claim that all workflows, datasets or controls are already live in production.
In progress
Evidence model, confidence policy, gated workflow and early example deliverables.
Not yet established
Live proprietary data pipeline, authenticated workspace, verified project inventory and capital-markets materials.
Core principles
Every advancement should be auditable.
These principles turn ATLAS away from a generic map product and toward a workflow system that can withstand challenge from landowners, operators, reviewers and investors.
01
Trace every material claim
Every material signal should carry a source, date, owner and confidence level before it influences a recommendation.
02
Separate reported facts from verified facts
Seller language, public listings and informal statements may enter the queue, but they should not be presented as confirmed infrastructure evidence.
03
Do not score past the evidence threshold
ATLAS should withhold readiness scoring when critical evidence is missing, contradictory or too stale to support a decision.
04
Keep unknowns visible
Unknowns should remain visible in the workflow instead of being averaged away inside a composite score.
05
Advance through explicit gates
A site advances only when the evidence needed for the next decision has been assembled and reviewed.
06
Preserve version history
Source changes, score changes and reviewer judgments should be recorded so later users can understand why a conclusion changed.
Evidence controls
A material signal must carry context.
Signals become useful only when ATLAS can show where they came from, how fresh they are and what decision they can legitimately support.
Control standard
Source traceability
URL or document reference, retrieval date and evidence owner.
Freshness
Date-aware evidence handling with explicit staleness warnings.
Confidence
Reported, documented or independently checked; never implied certainty.
Contradictions
Conflicting inputs stay visible until resolved by review.
Authority
Material claims should be tied to the party with authority to make them.
Decision use
Each signal should state what decision it can and cannot support.
Confidence policy
Confidence is a use policy, not a decoration.
ATLAS should downgrade what it claims when evidence is weak. Strong presentation should never compensate for low-confidence inputs.
Cannot support readiness claims, capacity assumptions or investment reliance.
Can support a preliminary memo, but still requires challenge before reliance.
Can inform advancement to the next gated step, not final engineering or legal conclusions.
Decision gates
The next step should be earned.
Each gate exists to answer a real development question. If the required evidence is absent, the site should pause instead of drifting forward.
Gate 1
Public signal screen
Listing or public signal exists, basic location and scale are known, ownership authority remains unconfirmed.
Queue for desktop screening or reject quickly.
Gate 2
Preliminary screen memo
Source-linked constraints, planning context, indicative access conditions and explicit unknowns are assembled.
Decide whether the site deserves deeper verification.
Gate 3
Qualified opportunity
Critical claims are documented, contradictions are logged and next external enquiries are defined.
Consider landowner alignment or structured development work.
Gate 4
Development mandate
Authority, workstreams, reviewer roles and commercial path are clear enough to justify spend.
Advance into controlled project development rather than speculative screening.
Scoring guardrails
No readiness score before minimum evidence exists for land, power, fiber and planning.
No conversion of proximity into capacity or entitlement.
No removal of negative signals simply because one category scores well.
No presentation of unverified commercial interest as demand proof.
No investor-facing summary without dated caveats where evidence remains incomplete.
Operating artifacts
Methodology note for external explanation
Source register with retrieval dates and owners
Constraint register with unresolved blockers
Decision memo with explicit next gate
Change log for score and evidence revisions
Important limitation. ATLAS concept scores, dashboards and public lead examples remain illustrative until backed by controlled evidence collection, dated sources and reviewer accountability. This page is meant to clarify that boundary, not blur it.