Skip to content

Insurance

Understand the handoffs, exceptions, and decisions that influence claims resolution.

One claim, traced across five parties

  1. Day 1Policyholder

    Claim submitted

  2. Day 2Claims handlerhandoff

    Coverage confirmed

  3. Day 6Claims handler

    Invoice requested

  4. Day 9Loss adjusterhandoff

    Site visit

  5. Day 12Loss adjuster

    Coverage re-checked

  6. Day 14Claims handlerhandoff

    Requested again

  7. Day 19Contractorhandoff

    Invoice → wrong record

  8. Day 20Loss adjusterhandoff

    Scope revised

  9. Day 21Senior reviewhandoff

    Over authority

  10. Day 21–26 · Senior review

    Waiting in the exception queue for 5 days

Handoffs
6
Coverage checks
2
Evidence requests
2
Days queued
5

Claims move between handlers, adjusters, outside firms and several systems. When one stalls or gets reviewed twice, it's hard to say why, and the customer just hears nothing.

Lavora connects your claims platform, policy admin, email, document portal, adjuster apps and payments. It rebuilds each claim's route: where it waited, and who decided what.

Missing evidence

Claims wait days for photos, invoices or proof of ownership, often asked for one at a time, sometimes twice.

44%of home claims chased

Manual handoffs

Referrals and adjuster reports leave the claims platform. Status lives in inboxes and gets re-keyed by hand.

6handoffs on claim C-30418

Duplicate review

Coverage confirmed by a handler, then checked again by another team with nothing new to go on.

38%re-checked by a second team

Opaque decisions

Exceptions queue for senior review, but the reasons and the evidence reviewers need are scattered.

5 daysqueued on claim C-30418

See how claims split between straight-through handling, evidence chases and exceptions. Switch claim type, or click any step to see who owns it and how long it takes.

Straight-through
476/ 1,000
Waited on evidence
440/ 1,000
Reached exception review
301/ 1,000
Settled by day 60
865/ 1,000
  1. Intake

  2. Documents

  3. Coverage

  4. Assessment

  5. Exception

  6. Approval

  7. Outcome

Every route 1,000 claims took. Tap any step for details.

waiting / rework human judgement

An average cycle time hides the claims stuck waiting on evidence, referrals and senior review.

Lavora compares paths, not averages, and shows what pushed each claim off the fast path.

Days from first notice to settlement

Each dot is one claim · bar = middle 50% · tick = median

Straight-through

Complete, desk-assessed, in authority

7.3d

baseline

Evidence chase

Waited on missing documents

18.5d

+11.2d

Coverage query

Referred to underwriting

21.4d

+14.0d

Adjuster + exception

External assessment and senior review

48.1d

+40.8d

Why claims reach exception review

Share of exception cases · recorded triggers

  • Reserve above handler authority46%
  • Adjuster scope change21%
  • Coverage interpretation14%
  • Complaint logged11%
  • Fraud indicator raised8%

Most exceptions are about authority, not doubt. A better-prepared file could shorten the queue without changing who decides.

Triage is repetitive, so it's tempting to hand it all to AI. But a misrouted complex claim can go unnoticed until the customer complains.

AI suggests a category and the handler confirms or overrides it. Lavora tracks overrides, samples quality, and sends high-value or uncertain claims straight to a person.

Home · reported via portal · photos attached

Water through kitchen ceiling

Suggested by triage assistant

Category
Escape of water, internal
Complexity
Standard
Confidence
High
Suggested route
Desk assessment
  • Description mentions burst pipe under upstairs bathroom
  • Photos show internal ceiling staining
  • Policy includes escape-of-water cover

Decision log · this queue

0 of 3 reviewed by a handler

Suggestions never move a claim on their own. Every confirmation and override is logged for quality review.

  • Override rate on AI suggestions

    observed

    11.8%

    Highest for disputed motor liability (24%), so suggestions there show as low confidence.

  • Sampled quality review

    policy

    5% of AI-assisted classifications re-checked weekly by a second handler. Agreement in the latest sample: 94% (observed).

  • modelled
    £15,000reserve and above

    Claims / week to second review

    70 of 420

    Reviewer time / week

    ≈ 53h

  • Always decided by people

    • Declines
    • Disputed liability
    • Fraud indicators
    • Vulnerable customers
    • Complex commercial losses

Design partners

Bring us the claims journey that frustrates your team most. We'll work out together where it slows down, and you'll help shape what Lavora builds next.