Swiggy Interview Questions & Answers 2026

Match orders to delivery partners based on: distance from restaurant, current load (each DP can carry 1-2 orders), partner rating, delivery time prediction. Use Hungarian algorithm or modified greedy with constraints. Real-time: handle DP going offline, traffic changes, batch orders for efficiency.

Components: order events (created, confirmed, prepared, picked, in-transit, delivered) on Kafka, WebSocket service for real-time client push, location updates from delivery partner app every 30s, ETA recalculation based on traffic. Handle: connection drops, order updates ordering, scale to millions of active orders.

Operations: getBalance, addMoney, debit, refund, transactionHistory. Tables: wallets, transactions (idempotency_key, type, amount, status). Handle concurrency (pessimistic lock or atomic SQL). Idempotent operations critical. Tests: race conditions, refund flows, partial failures.

15-30 min delivery via dark stores (1000-5000 SKUs each). Hub-and-spoke model. Demand prediction for SKU stocking. Dense network of dark stores (3-4 km radius coverage). Stock from manufacturers/distributors not retail stores. Tradeoff: lower assortment vs faster delivery. Higher operating cost than scheduled delivery.

Use two heaps: max-heap for lower half, min-heap for upper half. Maintain size difference of at most 1. Insert in appropriate heap, rebalance if needed. Median = top of larger heap, or average of tops if equal sizes. Time O(log n) per insert, O(1) for median.

Use HashMap + Doubly Linked List. HashMap maps key to DLL node for O(1) access. DLL maintains usage order - most recent at head. On get: move node to head. On put: if exists, update and move to head; if new and capacity full, remove tail node and add new at head. All operations O(1).

Swiggy is a leading food-tech with strong product culture and tech-first approach. Mention their innovations (SuperApp, Instamart, Genie, Minis), engineering blog (TechAtSwiggy), focus on operational excellence. Listed on NSE in 2024. Direct impact on hundreds of millions of monthly orders. Strong learning environment with multiple business units.

Use STAR method. Describe issue severity, symptoms, your investigation approach (logs, metrics, distributed tracing), root cause identified, fix deployed, post-mortem learnings. Swiggy's scale means debugging complex distributed systems is common.