Thirty-two industries. One exposure map.
GRIDINT's fragility engine scores every vertical it tracks — how exposed each industry's AI story is, and whether the payoff behind that story is demonstrated, emerging, or still just spend. The engine's scores, presented as the engine's scores.
Model outputs labeled as model outputs — method on file, never dressed as observed fact.
The map.
Each cell is a vertical. The wash is the engine's mean fragility — deeper means more exposed; the corner mark is its payoff status (● demonstrated · ◐ emerging · ○ spend). Unscored cells say so with a faint wash rather than a fake number. Tap any cell for its companies and its payoff evidence.
How this was made.
GRIDINT's fragility engine scores 76 tracked companies across the verticals; the vertical score is the mean of its tracked names — the reconciliation counts are in the record, not hidden.
Fragility and payoff are the engine's MODEL OUTPUTS. This page never presents them as observed facts — that labeling is the provenance rule this family runs on.
When the engine regenerates its series, this page's distill re-runs as one post-build pass — the exhibit tracks the engine, not a copy of it.
Not investment advice, and not a judgment on any company — an exposure map of an industry conversation, receipts on file.