You're viewing a live public demo on fictional data — nothing you change here is saved.

Error / rework · Investigating

Underwriting override rework

When an underwriter overrides the automated pricing rules, the rationale is written once as free text and never checked against actual claims outcomes, so the same mistakes recur.

Edit

Friction Priority

High (72/100) — which friction to investigate first, not how valuable the underlying information is (that's Information Advantage, a separate measure).

  • frequency: band 3/5 × weight 0.3 = 0.90
  • estimatedTimeCost: band 5/5 × weight 0.25 = 1.25
  • customerImpact: band 1/5 × weight 0.2 = 0.20
  • riskImpact: band 5/5 × weight 0.15 = 0.75
  • informationAvailability: band 5/5 × weight 0.1 = 0.50

Not yet incorporated (no data source in this system): business impact -- no field in this schema represents it yet; decision significance -- no field in this schema represents it yet; feasibility of intervention -- no field in this schema represents it yet.

Friction cost

Estimated time cost: 720 person-hours per period.

  • estimatedTimeCostHours = frequency x timePerOccurrence x peopleAffected -- a person-hours-per-period estimate, not a currency figure.
  • No error/delay/missed-opportunity cost was supplied -- additionalMonetaryCost is null, not zero, to distinguish 'not estimated' from 'estimated at nothing'.
  • customerImpact and riskImpact are not included in this calculation -- they feed Friction Priority (a separate, ranking-only signal), not the cost estimate itself.

Details

Department
Underwriting
Process stage
—
Owner
Chief Underwriter
Customer impact
Low
Risk impact
High
Evidence / notes
Illustrative estimate for this demonstration dataset, not a measured figure.

Information assets involved

Information Diagnosis

Friction → Information Diagnosis → Information Asset -- connects this friction to the information conditions behind it, where one exists.

  • Manual synthesis required

    Whether an override pattern has historically been risky is knowable in principle from Underwriting Decision Records, but nobody currently cross-references new overrides against that history in real time.

    Opportunity: Flag overrides matching historically poor-performing patterns at the point of decision, not months later in an audit.

    Assets: Underwriting Decision Records, Senior Underwriters' Risk Judgement Heuristics

AI Opportunities grounded in this friction

Grounded in the information assets involved above, plus this friction and any diagnosis of it, passed as context -- Friction is a discovery entry point in its own right, not only a filter on assets found some other way.

None yet.