Bellwether selection scored against the CMO.
Trial picks decide the value of the whole pool, yet selection is too often argued from intuition. When the Mass Tort edition opens, the bellwether sub-agent will read the case management order’s actual selection rules and score every plaintiff for representativeness against the inventory it sits in. It is designed to compose fact-sheet completeness, injury severity, causation density, jurisdiction risk, and defense exposure into a defensible ranking — so the steering committee walks into the selection conference with distribution math, not a hunch.
What MT is designed to do.
- 01
The sub-agent will parse the governing CMO — randomization pools, per-side strike counts, eligibility cutoffs — and score only against the protocol the court actually ordered, not a generic template.
- 02
Representativeness is designed to be measured as distance from the pool: how a candidate slate’s distribution of implant model, injury grade, and venue compares to the full inventory, surfaced as explicit math.
- 03
Justine will flag when an opposing slate skews — eight nominees clustered in the mildest injury tier or the most defense-friendly venues — and propose a counter-slate that mirrors the pool.
- 04
Every score will trace to the fact-sheet entries and record pages that produced it, so the steering committee can defend each pick to the special master rather than asserting a black-box ranking.
- 05
The selection argument is attorney work product: Justine reasons the slate and shows its reasoning; the steering-committee attorney designates the trial picks and signs the representativeness brief.
The AI reasons; the attorney decides.
JustineAI™ MT is coming next. This describes the workload it is built to carry. When it opens, founding-firm slots go to the mass tort attorneys who told us about their practice early.