What topics are tested in Razorpay SDE-3 interviews?

Razorpay SDE-3 interviews test: Distributed Locking, Domain-Driven Design, Error Contracts, Extensibility, Binary Indexed Tree, Tarjan's SCC, Advanced Graph Algorithms, Stream Processing.

Does Razorpay SDE-3 have a system design round?

Yes, Razorpay SDE-3 includes a system design round.

Razorpay

Q: How many rounds are there in Razorpay SDE-3 interviews?

Razorpay SDE-3 typically has 5–6 rounds. Razorpay SDE-3 interviews include dual system design rounds, a leadership discussion, and — as of 2024–2025 — an agentic coding round where the candidate drives an AI-assisted implementation session. The design rounds test settlement pipeline architecture, compliance constraints (RBI data residency, PCI-DSS), and SDK versioning for 200K+ merchants. Leadership questions are specific: cross-team alignment on competing priorities, zero-downtime migrations, and where to invest in payments infrastructure for the next three years.

Fintech · Bengaluru, India

Razorpay is India's leading payments infrastructure company, powering payment acceptance for 8M+ businesses.

Related Companies

Rounds: 5–6 rounds
Focus: Machine Coding + System Design + Leadership
System Design: Yes
Timeline: 4–5 weeks to offer

Role:

Rounds 1–2 are fully visible · Round 3+ unlocked with a free account

Machine Coding

Round 1 of 5LLD / OOD

90 min

Distributed LockingDomain-Driven DesignError ContractsExtensibility

Production-grade LLD where the bar is a codebase another team would be comfortable owning on Day 1. Recent SDE-3 problems include distributed lock managers, reconciliation engines, and in-memory search engines with ranking. At this level, Razorpay expects you to proactively define error contracts ('is this error retryable?'), plan for extensibility ('how do you add a new matching algorithm in 6 months?'), and write a comprehensive test plan without being asked.

Hard

What to expect

e.g., 'Build a Reconciliation Engine' — match payment events against bank settlement files, flag discrepancies, handle partial matches and clock skew between systems
e.g., 'Implement a Distributed Lock Manager' — Redis-based lease with TTL, heartbeat renewal, fencing tokens to prevent stale writes after lock expiry, quorum-based Redlock for split-brain, leader election
e.g., 'Design an In-Memory Document Search Engine' — inverted index, TF-IDF ranking, incremental index updates without full rebuild
Extensibility gate: 'How do you add a new reconciliation rule (e.g., match by bank reference instead of transaction ID) without rewriting the matching core?'
Error contracts: 'What does your lock acquisition API return when the lock is held? Is it retryable? How does a caller know the difference from a network timeout?'
Full test plan expected — concurrent lock acquisition, TTL expiry during operation, split-brain on Redis failover

Practice on KnowEase.AI

LLD Practice →

Coding Round

Round 2 of 5DSA / Coding

60 min

Binary Indexed TreeTarjan's SCCAdvanced Graph AlgorithmsStream Processing

Hard DSA connected to production payment infrastructure. At SDE-3, Razorpay frames algorithmic problems as optimizations you'd actually make in a payment system — the expectation is that you can translate an algorithmic insight into a system-level argument. Parallelization and memory pressure at scale are expected follow-ups.

Hard

What to expect

e.g., 'Design a data structure for real-time fraud scoring on a stream of payment events — support O(log n) range sum queries for sliding windows' — Binary Indexed Tree or segment tree
e.g., 'Find all strongly connected components in a merchant→acquirer→bank routing graph — these represent circular dependencies in settlement' — Tarjan's SCC
e.g., 'Given a large transaction log, find all merchants reachable from a flagged merchant within 3 hops' — BFS with depth bound, efficient adjacency representation
Parallelization follow-up: 'How would you partition this BFS to run across 10 machines handling 10M transactions/minute?'
Memory pressure: 'At what graph size does your adjacency matrix become impractical? What's your alternative and its trade-offs?'

Practice on KnowEase.AI

DSA Practice — Graphs →

System Design 1

Round 3 of 5System Design

60 min

Settlement ArchitectureEvent SourcingComplianceMulti-bank Integration

Design a core Razorpay infrastructure service — typically the settlement pipeline or transaction processing backbone. At SDE-3, you're expected to proactively raise compliance (RBI data residency, PCI-DSS), immutable audit trails, and failure recovery without waiting for prompts. An interviewer who has to ask 'what about compliance?' is already marking you down.

What to expect

e.g., 'Design Razorpay's Settlement Pipeline' — batch settlement to 50 banks, per-bank SLA tracking, reconciliation against bank files, exponential backoff retry for bank API failures
Audit trail: 'Every payment state transition must be immutable and auditable for RBI — how do you design this? Event sourcing vs. audit log vs. append-only ledger?'
Multi-bank abstraction: 'HDFC, ICICI, and Axis Bank APIs have completely different response formats and failure modes — how do you abstract this without leaking bank-specific logic into the settlement core?'
Compliance: 'RBI mandates payment data residency in India. How does this constrain your DR architecture? What regions can you fail over to?'
Bank outage: 'HDFC's settlement API is down for 2 hours during the 9 PM batch window. What does your pipeline do? How do merchants find out?'

Full details unlocked with a free account

API & SDK Design (LLD)

Round 4 of 5LLD / OOD

60 min

SDK DesignAPI VersioningSecurityError Taxonomy

Low-level design of an external-facing SDK or internal service interface. The focus is on versioning, backward compatibility, and error taxonomy — the problems that uniquely arise when 200K merchants depend on your API contract. Recent problems include the Razorpay Checkout SDK and notification delivery systems with configurable routing.

What to expect

e.g., 'Design the Razorpay Checkout SDK' — event hooks (on_success, on_failure, on_dismiss), iframe sandboxing, automatic retry with merchant-visible state, mobile vs. web differences
API versioning: 'v1 of your payment API is live with 200K merchants. You need to deprecate a field that 30K merchants depend on — what's your migration strategy and timeline?'
Error taxonomy: 'Distinguish retriable errors (network timeout, bank unavailable) from terminal ones (invalid card number, expired CVV) in your SDK's error contract — why does this matter to merchants?'
Security: 'How do you prevent merchant A from using merchant B's API key in your SDK? What if a merchant accidentally ships their key in a mobile app binary?'
Backward compatibility trap: most candidates design v2 correctly but forget that v1 clients will still call the same endpoints — interviewers check this explicitly

Full details unlocked with a free account

Leadership Round

Round 5 of 5Behavioral / HM

60 min

Technical LeadershipCompliance ThinkingMigration StrategyTechnical Vision

Technical leadership and ownership at SDE-3 scope — owning multi-team initiatives, setting reliability standards, and driving decisions across the payments stack. The questions are specific and expect specific answers; vague stories about 'influencing cross-functionally' do not pass at this level.

What to expect

Cross-team alignment: 'Tell me about a cross-team technical initiative you drove where teams had competing priorities. How did you get alignment without authority?'
Compliance vs. velocity: 'How do you balance RBI/PCI-DSS compliance requirements against engineering velocity? Give a concrete example where they conflicted.'
Zero-downtime migration: 'Describe a migration you led on a service that couldn't go offline — payments infrastructure specifically. What was your rollback plan?'
Technical vision: 'What is the most important technical investment Razorpay should make in its payments infrastructure over the next 3 years? Be specific — architecture, not platitudes.'

Full details unlocked with a free account

Questions — coming soon.

Company Guide — coming soon.

About Razorpay SDE-3 Interviews

Razorpay SDE-3 interviews include dual system design rounds, a leadership discussion, and — as of 2024–2025 — an agentic coding round where the candidate drives an AI-assisted implementation session. The design rounds test settlement pipeline architecture, compliance constraints (RBI data residency, PCI-DSS), and SDK versioning for 200K+ merchants. Leadership questions are specific: cross-team alignment on competing priorities, zero-downtime migrations, and where to invest in payments infrastructure for the next three years.

See interview loops for other companies:

Microsoft Amazon Uber Flipkart Walmart Coupang Rippling Rubrik Target

Start your Razorpay SDE-3 prep plan

Practice the exact topics tested in each round — DSA, system design, SQL, and LLD.

Start free →