An independent verdict on your subsurface model — signed either way.

In-sample fit is not evidence.

Interpretation platforms tie a velocity model to your wells, and the tie looks good because that is what tying means. We withhold wells, test the model on those, and freeze the pass mark before we see your data.

The problem

Procurement asks for proof; vendors sell processing.

Imaging contractors compete on turnaround and GPU count. Interpretation platforms compete on cycle time. Neither sells what subsurface leadership actually needs before a depth model enters reserves: an independent check that names REFUSE when the acquisition cannot support the question — and signs the receipt when it can.

What you get

A signed assessment, not a volume handoff.

1. 01

   Four acquisition gates, cheapest first

   Instrument clearance, blind-well illumination, target depth reachability, and spatially blocked holdout — each with a measured floor. A gate REFUSE stops spend before inversion runs.

2. 02

   Blind-well mistie against a measured truth floor

   Depth error at withheld markers, decided with well-resampled bootstrap intervals. REFUSE means the method failed the bar — not that we ran rigorous QC and hoped.

3. 03

   Cryptographic receipt

   Every verdict ships with bundle digests and reproduce commands your team or auditor can run without our stack. PROMOTE and REFUSE use the same chain.

Why believe us

REFUSE is a first-class outcome.

Most vendors cannot afford to name failure — their revenue depends on the answer being yes. Every assessment ships PROMOTE or REFUSE with the same cryptographic receipt your auditor can replay offline.