Company AI fingerprint

Two companies, six capability axes, one chart. Where each polygon bulges tells you what that company is actually building. The asymmetries — not the absolute values — are the story.

Company A:
Company B:
Foundation trainingFoundation training NVIDIA: 21% (109/530) Scale AI: 35% (35/101)21% / 35%Inference infraInference infra NVIDIA: 79% (420/530) Scale AI: 46% (46/101)79% / 46%Agent platformAgent platform NVIDIA: 32% (170/530) Scale AI: 46% (46/101)32% / 46%Eval & safetyEval & safety NVIDIA: 27% (141/530) Scale AI: 59% (60/101)27% / 59%RAG / retrievalRAG / retrieval NVIDIA: 12% (64/530) Scale AI: 17% (17/101)12% / 17%Robotics + multimodalRobotics + multimodal NVIDIA: 25% (134/530) Scale AI: 19% (19/101)25% / 19%Research depthResearch depth NVIDIA: 12% (64/530) Scale AI: 28% (28/101)12% / 28%NVIDIA · Foundation training: 21% (109/530)NVIDIA · Inference infra: 79% (420/530)NVIDIA · Agent platform: 32% (170/530)NVIDIA · Eval & safety: 27% (141/530)NVIDIA · RAG / retrieval: 12% (64/530)NVIDIA · Robotics + multimodal: 25% (134/530)NVIDIA · Research depth: 12% (64/530)Scale AI · Foundation training: 35% (35/101)Scale AI · Inference infra: 46% (46/101)Scale AI · Agent platform: 46% (46/101)Scale AI · Eval & safety: 59% (60/101)Scale AI · RAG / retrieval: 17% (17/101)Scale AI · Robotics + multimodal: 19% (19/101)Scale AI · Research depth: 28% (28/101)NVIDIA· 530 AI roles, 8 regulatedScale AI· 101 AI roles, 9 regulated
Each axis = share of company’s tagged AI roles touching that capability cluster. Outer ring = 100%, inner ring = 25%. Hover any vertex for the underlying count.

Suggested comparisons