CONTEXT(CLONE)
Sign in

Prepare a coverage brief with clauses for and against

For: Claims handler or coverage specialist preparing a position on a complex commercial loss

Pattern: Cross-examinationRuns todayDesigned for 4 to 32 agents
Get early accessSign in to use

The pain today

The schedule, wording, endorsements and loss report each run long. Handlers read them under time pressure, and a missed endorsement or condition is found only when the position is challenged.

The ask

“I attached the policy schedule, wording, endorsements and the loss adjuster's report for one claim. Which clauses are engaged by the reported facts, for and against cover? Quote each clause and the fact it relies on, and show me where the readings disagree.”

Plain words, as you would say it to a colleague. Edit it to fit your case before you send it.

What you attach or connect

  • Policy schedule
  • Policy wording
  • Endorsements
  • Loss adjuster's report
  • Notification correspondence

The unit of work

One worker task per one policy document or report section per reader.

Why a swarm fits

Each reader needs one document and the list of reported facts. Two families read independently so a shared misreading of a clause is less likely, and disagreements reach the handler instead of being smoothed over.

Not for

Deciding the claim. It lays out clauses and facts; it gives no coverage decision, reserve or settlement figure.

The decision tree

7 typed decisions, each with an action for every answer

At fixed moments in a run, the engine puts one narrow question to a decision model. The decision model never writes text: it answers yes or no with a probability, picks from listed options, or gives a score, about a small slice of the material. The engine then does exactly what this tree says, which is what makes the run auditable. The thresholds are the template's design values, not measured results.

  1. Planner, while planning

    Scope checkYes or no, with a probability

    Before work starts on a unit

    Do the dates and policy number on this endorsement or wording show it formed part of the policy for the period stated in the schedule?

    Sees only: The schedule's period and policy number, and one document's header and effective date

    Why: Stops a brief from quoting an endorsement that was never part of this policy.

    • Yes: 0.60 or higherthenAccept
    • Unsure: 0.30 up to 0.60thenEscalate to a strong model
    • No: below 0.30thenSkip this unit

  2. After workers, the judge checks

    Evidence checkYes or no, with a probability

    After a worker answers

    Does the loss adjuster's report state this fact in the quoted sentence, rather than the reader having inferred it?

    Sees only: One listed fact and the report sentence quoted for it

    Why: The brief may only use facts someone accountable wrote down.

    • Yes: 0.85 or higherthenAccept
    • Unsure: 0.50 up to 0.85thenMark unresolved
    • No: below 0.50thenReject and retry

  3. Evidence checkA choice among options

    After a worker answers

    Do the words of the quoted clause reach the reported fact it has been matched to, on a plain reading of the clause and its definitions?

    Sees only: One quoted clause, its definitions and one reported fact

    Why: Separates clauses that plainly apply from those whose application is the very thing in dispute.

    • Clause reaches the factthenAccept
    • Clause does not reach itthenReject and retry
    • Turns on interpretationthenMark unresolved

  4. Reconciler, while merging

    Conflict checkYes or no, with a probability

    While reconciling

    Do the two independent readings list different clauses as engaged by the same reported fact?

    Sees only: Both readers' clause lists for one reported fact

    Why: A clause only one family noticed is either a catch or an error; both are worth a stronger look.

    • Yes: 0.70 or higherthenEscalate to a strong model
    • Unsure: 0.30 up to 0.70thenEscalate to a strong model
    • No: below 0.30thenAccept

  5. Run control, between rounds

    Retry or stopYes or no, with a probability

    After a rejection or low confidence

    Can the difference between the two readings be settled from the policy text alone, without weighing evidence about the loss?

    Sees only: The two readings, the clauses they cite and the matched fact

    Why: Text questions can be retried; questions of evidence stay open for the handler.

    • Yes: 0.85 or higherthenEscalate to a strong model
    • Unsure: 0.50 up to 0.85thenMark unresolved
    • No: below 0.50thenMark unresolved

  6. After workers, the judge checks

    Evidence checkYes or no, with a probability

    After a worker answers

    Is the drafted brief free of any statement that the loss is covered, not covered, or worth a stated amount?

    Sees only: The drafted brief

    Why: The run prepares evidence; a conclusion on the claim must never slip into the text.

    • Yes: 0.90 or higherthenAccept
    • Unsure: 0.60 up to 0.90thenReject and retry
    • No: below 0.60thenReject and retry

  7. Accountable person, before anything is settled

    Person decidesYes or no, with a probability

    Before anything is reported as settled

    Does this brief concern a claim on which a coverage position has yet to be taken by a person?

    Sees only: The brief's header and its list of open points

    Why: Every brief goes to the named claims handler, who alone decides and records the position.

    Accountable: The named claims handler makes and records the coverage decision; disputed readings go to coverage counsel.

    • Yes: 0.10 or higherthenAsk a person
    • Unsure: 0.02 up to 0.10thenAsk a person
    • No: below 0.02thenAsk a person

The fleet: who does what

Model tiers by role, not brands: you choose the models. Strong reasoning models plan and reconcile, small fast models do the wide work, and the judge is a decision model from a different family, so it does not share the workers' blind spots.

  1. Planner

    A strong reasoning model lists the reported facts and the questions each policy document must answer.

    Decisions here:1. Scope check

  2. Workers

    Two sets of small fast workers from different families read each document independently and quote engaged clauses.

    Designed for 4 to 32 agents, one worker task per one policy document or report section per reader. Each worker receives only its own unit.

  3. Judge, from a different model family

    A decision model from a third family settles only disagreements the text resolves, and keeps the rest open.

    Decisions here:2. Evidence check3. Evidence check6. Evidence check

  4. Reconciler

    A strong reasoning model lays out clauses for and against cover, with no conclusion on the claim.

    Decisions here:4. Conflict check5. Retry or stop

  5. Accountable person

    The named claims handler makes and records the coverage decision; disputed readings go to coverage counsel.

    Decisions here:7. Person decides

Checked before anything is accepted

  • Every engaged clause is quoted with the reported fact it is matched to
  • Facts are taken only from the supplied report, never assumed
  • Endorsements are checked for overriding each quoted clause
  • The brief is checked to contain no coverage decision or amount

What comes back

  • Clauses pointing towards cover, with quotes and facts
  • Clauses pointing against cover, with quotes and facts
  • Conditions and time limits that apply
  • Open disagreements and facts the report does not establish

What to measure

  • Clauses the handler found relevant that the brief missed
  • Handler time to a documented coverage position
  • Share of open disagreements counsel considers genuine
  • Cost per brief compared with one strong model alone

Names of measures only. No result is claimed for this template.

Templates open in the workspace chat with the ask filled in. Nothing runs until you send it.

Get early accessSign in to use

More in Insurance and claims

Map one exposure across the whole wording portfolio

For: Wordings manager or portfolio underwriter responsible for product wordings

When a new exposure appears, nobody knows which wordings, endorsements and legacy versions cover it, exclude it or say nothing.

Pattern: Map, verify, reduceNeeds scale6 decisionsDesigned for 40 to 800 agents

Watch wording changes and recheck only what depends on them

For: Product governance lead or wordings manager at an insurer or managing agent

A clause edit in a base wording silently affects endorsements, customer summaries, claims guidance and broker templates built on it.

Pattern: WatchtowerNeeds a connector5 decisionsDesigned for 20 to 500 agents

Test a fraud pattern hypothesis before it is acted on

For: Special investigations lead examining a suspected organised pattern across referred files

A suspected pattern, such as the same repairer invoices across many files, is persuasive once stated.

Pattern: Adversarial reviewNeeds live models6 decisionsDesigned for 10 to 200 agents