Navdeep Singh, known as NeetCode, built one of the most popular coding interview platforms on the internet after leaving Google. In this Pragmatic Engineer episode, he argues that leetcode-style interviews persist not because they work, but because they scale. Large companies need to train hundreds of interviewers fast, and algorithm puzzles are easy to standardize. The process has nothing to do with predicting job performance.
The more important argument Singh makes is about learning hard things as a compounding investment. He left Amazon after two months, spent time at Google, then bet on himself full-time. His case is that deep work on difficult problems builds the judgment that survives tool changes, including AI. Systems thinking and domain expertise are not being automated. Grinding LeetCode, done right, is practice for that kind of depth, not just a hiring filter.
The episode earns a full read for what Singh says about the decision to leave a stable engineering career, what Google actually taught him versus what Amazon did not, and his specific take on why AI raises the floor for engineers without eliminating the ceiling. The sponsors include Antithesis, used by Jane Street and Fly.io for hostile-environment testing, Sentry with its Seer AI debug agent, and Google Cloud Run with automatic zonal failover.
[READ ORIGINAL →]