FAANG : The Journey | Post 4 - Interview Experience

This Post is just about the type of questions asked at a FAANG company rather than an experience of story telling.

Amazon L6 | Google L5 | June 2020

Amazon:

• Online assessment: 2 problems in 90 mins. First problem about frequency calculation, ranking and returning top K. Second problem on sorting based on a custom criteria.
• Virtual onsite round 1: First 25 mins LP. Then coding: Graph path finding between given nodes (hard). Follow up questions like why dfs/bfs, backtracking vs other approaches.
• round 2: First 20 mins LP. Then coding: problem with geometry (medium). Like solving line/circle equations to determine points on the canvas. Follow up question on how to handle the problem on a very large scale say the number of points don't fit in one machine's RAM.
• round 3: First 25 mins LP. Then coding: problem on graph traversal (medium). Follow up questions on bfs/dfs, recursive vs iterative solution.
• round 4: First 25 mins LP. Then system design: it was related to designing a system to handle large streaming data and a dashboard with various aggregations. In depth discussions on type of database to use (tradeoffs), scaling, cache, availability and resiliency of the system etc.

Google

• Phone screen: coding - resource allocation problem. Given M resources with different capacities and N consumers, allocate N to M optimally.
• Virtual onsite 1: Coding - Graph find common connections at nth depth (medium/easy). The problem statement was very open/vague and the interviewer asked me to think about the solution as a product manager. So I think apart from code, clarifying and coming up with an approach was the key here.
• Virtual onsite 2: Coding - Problem based on Tries and PriorityQueue (medium). Given the data, the system required to answer two type of queries, one based on string prefix and another based on a priority. The interviwer asked to implement Tries in addition to the actual problem.
• Virtual onsite 3: Coding - code to sovle sudoku based on some conditions (Hard). I could not solve this without hints and could not finish the coding in time.
• Virtual onsite 4: Leadership - usual leadership & behavioural questions.
• Virual onsite 5: System Design - Usual system design for Google scale image storage. But expect to go deep with numbers (throughput, latency, storage...). The interviewer went very deep with how to decide the number of servers, database throughput capacity, cache memory and so on. A lot of discussion on tradeoffs.

Reference :https://leetcode.com/discuss/interview-experience/714973/nda-amazon-l6-google-l5-june-2020-offer-offer