1. Supply your comparison
2. Inspect ranking and trade-offs
A non-dominated option has no other eligible option at least as good on every rating and strictly better on one. It can still rank below another option under your selected weights.
Review the evidence and uncertainty yourself. Eligibility is a supplied threshold check, not permission to act. Equal scores use ASCII identifier order; excluded candidates have no rank.
Inspect complete report JSON
Reproduce the comparison
The original Python demo reuses the repository’s Pareto comparison engine. This browser uses parity-tested integer arithmetic and the same canonical input hash. Both preserve supplied evidence and make no external calls.
Files remain in this tab. Export to keep your work. Reopening a report ignores its saved result and recomputes from the input. Notebook evidence is also supported within the 1 MiB browser limit. For larger notebook envelopes, reopen the exported cross-industry-input.json instead.
Original runtime, schema and launch guide