Real Interview Experiences

Phone screen covered sliding window and two pointers. Onsite had 4 coding + 1 Googleyness round. Questions included Trapping Rain Water (modified), Word Break II, and a graph shortest-path variant. HC approved after a borderline decision. Tips: focus on clean code and verbalize edge cases throughout.

OA had 2 LC-medium problems (arrays + hashmaps). Interviews were 2 coding + 1 HLD + 1 behavioral. LPs are taken very seriously — prepare 8+ STAR stories. Coding problems were LRU Cache and Meeting Rooms II. HLD was "Design Amazon Cart service at scale". Received offer in 5 business days.

Meta's process was very structured. 2 coding rounds (speed is key — both medium/hard), 1 system design, 1 behavioral. Coding had Clone Graph and a Trie variant. System design: design Instagram newsfeed. The behavioral was conversational, not STAR-heavy. Recruiter communicated quickly throughout.

Process was smooth. Coding rounds focused on trees and linked lists — Binary Tree Level Order Traversal and Merge K Sorted Lists came up. System design was design a notification system. Microsoft values "growth mindset" in behavioral. Expect Cracking the Coding Interview style questions rather than tricky graph problems.

Stripe's process is uniquely structured — take-home project (1 week), then 3 technical rounds. Take-home involved building a simplified payment API. Rounds covered API design, debugging, and a distributed systems deep-dive on rate limiting. Very engineer-friendly culture. Focus on code quality not just correctness.

Algorithm-heavy process — expect true hard-level problems. Minimum steps to reduce array, interval scheduling maximization. Also had 1 system design (design TikTok FYP algorithm at a high level). Move fast on coding — they care about time per problem. Received feedback that my initial approach was slow.

Uber focuses heavily on maps/geo problems. Expect shortest path, spatial indexing, and matching algorithms. System design: design surge pricing system. Behavioral was one round focused on cross-functional collaboration. The team fit call was a real differentiator — be genuinely curious about the product.

Netflix's bar is extremely high. My system design (content delivery + CDN) was flagged for not going deep enough on edge-caching strategies. Coding was fine — 2 medium problems on arrays and DP. The architecture discussion expected me to have deep knowledge of streaming protocols (HLS, DASH). Feedback was constructive.

Very product-focused interviews. One round was literally pair programming on a real Airbnb feature. Coding included calendar booking overlap detection (Meeting Rooms variant) and sorting. System design: design the Airbnb search and ranking system. The product sense round surprised me — they asked "how would you improve our checkout flow?"

Very engineering-heavy. Expect distributed systems questions throughout. 2 coding rounds (one had a problem on consistent hashing), 1 open-source contribution discussion, 1 HLD (design Delta Lake compaction), 1 behavioral. They loved that I had experience with Spark. Deep technical conversations rather than pure algorithm grinding.

Shopify's process is fast and practical. Take-home Ruby on Rails challenge (build a mini Shopify checkout) then 2 technical interviews. One was system design (flash sale system), one was debugging a real-ish codebase. Very collaborative — interviewers feel like partners not judges. Strong async/remote culture.

Spotify uses their "Squad Model" heavily in behavioral. Technical rounds covered graph algorithms (playlist recommendation graph traversal), system design (music recommendation at 600M MAU scale), and a home assignment building a music search index. Great culture interview — they value autonomy and ownership.

Heavily focused on distributed systems and observability. Coding had time-series manipulation and a streaming aggregation problem. System design: design a distributed metrics pipeline (think Prometheus at scale). Technical depth expected — know your data structures for time-series (LSM trees, SkipLists).

LinkedIn's process was well-organized. Coding problems focused on graphs (BFS/DFS for connection degree, LRU cache). System design: design LinkedIn's feed ranking system. Behavioral emphasized "member focus" — they want you to talk about users. Strong values alignment expected in behavioral round.

Palantir's process is uniquely long — expect a multi-day hackathon-style final round. Earlier rounds covered algorithm depth (graph entity resolution, interval merging at scale) plus a decomposition exercise. They deeply value ownership and first-principles thinking. Be prepared to discuss trade-offs thoroughly.

OpenAI's bar is extremely high across both ML systems and software engineering. Coding covered distributed task scheduling and a tokenizer implementation. System design: design the OpenAI API inference infrastructure. They deeply care about mission alignment — research discussions are genuine, not performative.