← Back to case study

Assurance Fit

An eligibility-and-relevance sandbox for the room that has to sign off on an assurance benefit before it ships.

Prototype by Ruchi Parijat
Product concept · not affiliated with any brand

Who this is for: Product, Legal/Compliance, Finance, Brand and an InsurTech partner, in the room where eligibility rules and member-relevance logic get argued over before launch. It is not a member-facing screen.

Before an assurance benefit reaches a member, this group has to agree on two different things: who's even allowed to have it (a policy fact Legal and Finance sign off on), and what's actually worth showing this member first (a judgment call Product owns). Those two questions get debated together constantly and should never be answered by the same logic. This sandbox keeps them apart, on purpose, so each stakeholder can push on the piece they own.

1 Load a scenario — or set your own booking context to stress-test an edge case someone in the room raises.
2 Check eligibility — the rules panel Legal and Finance would actually review and sign off on.
3 Check the recommendation — the relevance call Product owns, with its reasoning exposed so it can be challenged, not just trusted.

Scenarios worth arguing about in the room

1 Booking context

Trip details

$640
21d

Blue members see standard eligibility and pricing across all assurance products.

2What's eligible (rules)

3 What's recommended, and why

—

—

—
relevance score
    ! Risk signal — raises urgency + Positive signal — supports relevance ~ Contextual signal — tier or timing nuance

    How this works: eligibility is deterministic — brand, tier, rate type and trip value decide what a member can buy. The recommendation on top is a weighted relevance score across cancellation risk, member tenure signal and time-to-check-in, the same split between rules and judgment used across the personalization work in the resume this prototype accompanies.

    For the room

    Open questions this logic surfaces

    A sandbox like this earns its keep by making stakeholders argue about the right things. These are the judgment calls baked into the current version — every one of them is a live debate, not a settled answer.