Business of AI · Gallery

Vendor Evaluation & Risk Monitor

SIMULATEDVerified Jul 2, 2026

The scorecard picks a winner — but move the weights and watch it change. That fragility is why the risk view matters: concentration and exit cost tell you what it costs to be wrong.

Same instrument · three industries — pick a use-case to reconfigure the run

Weights

Capability25%

Security20%

Roadmap15%

Low lock-in15%

Support10%

Price / value15%

Open-source-backed

Top pick

Portable, cheaper, thinner support.

Specialist

Best-in-class capability, narrower surface.

Hyperscaler

Broad platform, deep integration, real lock-in.

Top pick: Open-source-backed — exit cost $90k

Nudge two or three weights and the ranking can flip — a "winner" that survives only one weighting isn't a decision, it's a preference. Pair the pick with its concentration and exit cost before you sign.

Steering-committee takeaway: The scorecard picks the vendor; the concentration view tells you what it costs to be wrong.

How this is built

Weighted score = Σ(weight × criterion score) ÷ Σ(weights), so weights are relative and always normalize to 100%. Lock-in and price are scored so higher = better (less lock-in, better value).

Risk view pairs each vendor with concentration (spend share if primary), renewal window, and exit cost (tracks lock-in). Stack: Next.js (static) + shared design system; client-side.

Limitations: vendor scores are archetype illustrations, not a live evaluation; real scoring needs POCs and reference checks. It exposes ranking fragility and exit exposure, not a procurement verdict.