CRED

CRED interview process (2026)

Blog / CRED interview process (2026)

CRED interview reports describe several ways to assess software engineering skills: algorithm problems, in-memory machine coding, project discussion and system design. The format varies by role and report. Confirm your own stages, allowed tools and deadlines with the recruiter. This guide was reviewed in November 2026 and includes explicitly dated older examples.The linked exercises are TechPrep practice adaptations of reported tasks. Their method signatures, constraints, tie-breakers and grading conventions make the exercises reproducible. They are not transcripts or a forecast of your interview. For example, a December 2025 SDE1 report describes extending an existing metrics codebase, while a December 2025 SDE2 account, published in January 2026 describes flight booking and a prototype assignment.

Confirm the format: Ask which stages apply to your role, whether coding is live or take-home, what tools you may use, and how to submit your work. The public reports reviewed here do not establish one fixed company-wide sequence or hiring timeline.
Machine coding and implementation: Reported tasks include payment-instrument recommendation in March 2024 and timestamp-range metrics in December 2025. A November 2021 contact-directory report asked for in-memory add/update and search behavior. Treat its date and exact scope as historical evidence.
Algorithms and problem solving: A November 2024 backend internship report names Accounts Merge. November 2021 iOS and December 2021 SDE2 reports supply the older cache, stack, rotated-array and account-settlement examples in this collection.
System design: The reports include flight booking with seat locks and vendor coordination, a driver-priced ride marketplace, and a historical referral leaderboard. State the requirements and work through the failure cases that follow from that scenario.
Project discussion or prototype: The December 2025 SDE2 candidate reports a 24-hour prototype with AI tools allowed. That permission and deadline belong to that account. Be ready to explain your own implementation choices, testing and limitations, and confirm the policy for your assessment.

Use the following study areas to organize your preparation. The collection contains five algorithm exercises, six low-level design exercises and one high-level design walkthrough.

Algorithms and data structures: Practice merging accounts, caches, stacks, binary search and minimum debt settlement.
Machine coding and low-level design: Implement precise state transitions, lookups, updates and deterministic results.
System design and coverage gaps: Study a leaderboard, then work through the reported booking, ride-pricing and ingestion variants.
Projects and behavioral preparation: Prepare concrete examples of implementation choices, testing and engineering tradeoffs.

1. Algorithms and data structures

Start with Accounts Merge, which is explicitly named in the 2024 backend internship report. Explain why email overlap connects whole groups and how your approach handles transitive merges.For historical practice, the November 2021 iOS account names LRU Cache and Min Stack. Explain the invariants behind constant-time operations, including updates to an existing cache key.The December 2021 SDE2 report describes Search in Rotated Sorted Array and Minimum Transactions to Settle Debt. Practice identifying the sorted half in binary search and reducing debt settlement to net balances. These dated examples support preparation but do not establish how frequently CRED asks them today.

2. Machine coding and low-level design

The 2024 payment screenshots describe eligible methods, amount limits, device UPI support and ordering by method priority and relevance. In Payment Instrument Recommender, test policy replacement, inclusive amount limits, tied relevance and instrument updates. The practice policies are configurable exercise data.The November 2021 contact task supports add/update, exact and prefix searches, case-insensitive name matching and result counts. Use Contact Directory to practice removing stale index entries when a contact changes.A December 2025 report asks for latency, throughput and error-rate metrics within a timestamp range, with p90 as an extension. Event Metrics gives that task explicit interval endpoints, units and empty-result rules. Its graded API and acceptance of out-of-order events are practice conventions.For storage, the CRED section of a September 2021 multi-company report names a simple key-value store. Start with In-Memory Key-Value CRUD Store. The October 2021 wide-column report includes secondary indexing, and a separate December 2021 backend report describes primary and secondary indexes. Indexed Wide-Column Store combines sparse rows with maintained equality indexes. Its string-only cell model is an explicit practice choice. The typed-key variation reported elsewhere remains a discussion follow-up.The 2024 internship account includes file insertion, deletion, moving and search, with collision, wildcard and permission follow-ups. File Manager grades subtree moves and exact/wildcard search in a serial tree. Discuss access control and concurrent changes separately. Passing its tests does not establish thread safety.For each exercise, state an invariant, implement the smallest coherent model and test a mutation that could invalidate earlier results. This is general preparation advice.

3. System design and coverage gaps

Use Top-K Leaderboard for ranking, score updates and ingestion tradeoffs. The December 2021 referral-contest report also mentions change data capture. Work through where referral events originate and how updates reach the ranking view. The generic walkthrough does not fully specify referral-domain rules or that candidate’s CDC discussion.Several reported scenarios need additional practice beyond the linked walkthrough. The December 2025 flight-booking account describes one-way and round-trip searches, connections, per-seat pricing, seat locking and cross-vendor double-booking prevention. Trace a reservation from selection through expiry or confirmation.The December 2025 InDrive variant lets drivers list prices and prevents riders from counteroffering. Explore how discovery, offer visibility and matching behave under that rule.A mobile-engineering report updated in October 2024 names a mobile analytics SDK. Its interview date is not established by the page-update date. The October 2021 account describes Splunk-like log management across time periods. These SDK and log-manager scenarios remain coverage gaps in this collection. File permissions, concurrent updates and typed indexed values also require separate discussion. The roadmap makes no claim to grade these follow-ups.

4. Projects and behavioral preparation

Prepare examples from your own work: a design decision you changed after testing, an incident you investigated, a tradeoff you explained to a teammate, and a feature whose requirements changed. Describe your responsibility, evidence, decision and outcome. These are general practice prompts, not questions attributed to CRED.For a prototype review, be ready to run a small demonstration, explain the tests and identify unfinished behavior. If AI tools are permitted, explain how you checked their output and where you made independent decisions.Use the Behavioral Playbook to organize your examples. The roadmap includes a general behavioral practice set alongside the source-backed technical exercises.

Conclusion

Begin with the CRED roadmap, then choose a machine-coding exercise and verify its state changes with focused tests. Revisit the reported design scenarios separately, especially booking coordination, driver-set prices and concurrency. Confirm your interview format with the recruiter and use the source dates to distinguish historical examples from newer reports.