Scenarios interview questions for insurance roles
89 questions from core to advanced: concepts, calculations, whiteboard diagrams, SQL, scenarios and behavioural. Answer out loud first, then reveal the model answer and example. Mark each one practised when you can answer it without looking.
Practised 0 of 89
The rating engine shows a premium ₹150 lower than the actuarial sheet. How do you investigate?
- Compare line by line: OD, loadings, discounts, NCB, TP, add-ons, GST.
- Check rounding (per line vs total), effective dates, rate versions, discount caps.
- Reproduce with one test case, document root cause, add a regression test.
Design an AI-assisted motor claims flow. Where is the human in the loop?
- Customer uploads photos → AI damage detection and repair estimate.
- Auto-approve under a value and confidence threshold; refer otherwise.
- Fraud score and policy checks run in parallel.
- Human surveyor for high value, low confidence, or fraud flags; every decision logged for audit.
How would you gather requirements for a Bima Sugam integration?
- Study the published API and standard forms; identify insurer-side gaps.
- Workshops with product, IT, compliance, finance.
- Define product catalogue mapping, quote and issuance APIs, payment and policy status callbacks, e-insurance account linkage.
- Plan sandbox testing and reconciliation.
A customer's renewal shows 0% NCB but they claim 3 claim-free years. What do you check?
- Gap between expiry and renewal (over 90 days → lost).
- Any claim on the expiring policy, ownership change, previous insurer confirmation.
- Data migration or rollover capture errors; NCB declaration and verification process.
Business wants straight-through issuance for health up to ₹25 lakh SI. How do you respond?
- Clarify the goal (conversion vs risk).
- Bring loss data: higher SI drives anti-selection.
- Propose tiered STP with digital medical questionnaires, tele-underwriting, data-based risk scores, and monitor claim ratios after launch.
The TP revision is notified with 15 days' notice. What is your plan?
- Load new slabs as new effective-dated rows; keep old rows for back-dated quotes.
- Check long-term TP and bundled variants, EV and hybrid logic, commission impact.
- Update portals, aggregator APIs and premium tables in communications.
- Regression on all business types; go-live checklist with rollback.
Production defect: for 3 weeks, 2,000 car policies were issued with NCB applied to third-party premium. What do you do?
- Contain: stop the defect (hotfix or rule switch) and confirm no new policies are affected.
- Quantify: list the policies and the TP shortfall per policy.
- Decide with compliance and finance: TP is a statutory premium, so the shortfall must be handled; agree whether the insurer absorbs it or collects it.
- Fix root cause, add regression tests, and report what happened.
ExampleRoot cause: the NCB rule was applied to 'subtotal' after a refactor instead of 'OD before NCB'. New test: NCB line must never change when only TP inputs change.Estimate how many private car policies come up for renewal in India each month.
- State the method before the numbers: interviewers judge structure, not the exact answer.
- Cars on the road × share with active insurance ÷ 12, since most policies are annual.
- Adjust for new cars on 3-year TP (their OD still renews yearly) and policies that lapse.
- Sense-check against a source if one is available.
ExampleAssume 5 crore cars and 60% insured: 3 crore policies ÷ 12 ≈ 25 lakh renewals a month. These are assumptions for the method, not market data.Claim turnaround in health has slipped from 2 hours to 5 hours for discharge. How do you find the cause?
- Break the time into stages: hospital request, TPA receipt, document check, decision, communication.
- Use timestamps from the claim system to find which stage grew.
- Check volume, staffing, rule changes and integration failures for that period.
- Remember the 3-hour rule: late approvals cost the insurer money, so quantify that too.