A dispatch plan built for the expected day fails on the day that isn't expected. FENRIR builds a plan that survives the bad case and still leaves room to react.
It doesn't lock into one rigid schedule. FENRIR fixes only what it must commit now and keeps adjustable room to react as conditions change through the day. Two-stage adjustable-robust optimization with linear decision rules.
Most tools assume they know the probability of each outcome. FENRIR stays robust even when that assumption is off. Wasserstein distributionally-robust optimization.
It weights the worst outcomes more heavily than the typical one, so a rare spike doesn't set the month back. Spectral (tail-weighted) risk in the objective.
The worst-case scenarios FENRIR builds become the stress tests STRIX runs and the surge signal TARSIS guards against. Shares its adversarial scenario set with STRIX and TARSIS.
During onboarding we validate the method on your own data — a 12-month walk-forward (non-overlapping folds, five benchmarks, Diebold–Mariano significance) — so you can see how it performs before you rely on it.
These are the real methods behind FENRIR, each in one plain line. We name the method as a credibility signal — we don't publish the recipe.
Plans recourse that adapts as the day unfolds, instead of committing to one rigid schedule.
Stays robust even when the forecast's probabilities are wrong, not just to a fixed scenario set.
Weights the bad tail more heavily than the average outcome.