Faced an interview for a front-end position at the company, the algorithm difficulty was unexpectedly high, and I failed. The interview standards seemed puzzling, as the content was very diverse for a front-end role. I have now let go of my fixation on Google, no need to make things difficult for myself.
Process:
Initial chat with HR, followed by a virtual onsite three weeks later. No initial interview stage, moved directly to onsite. Onsite consisted of five rounds: four technical rounds (originally supposed to be three algorithm rounds and one front-end round), and one round of behavioral questions.
First Round: A variant of a “LeetCode” question, more challenging than expected. The interviewer asked C++ questions, which were difficult from the start. Tried using the calculator 3 solution for some cases but struggled.
Second Round: Tasked with creating an API to fetch data from a database and return it to the front end, as well as how to enhance concurrency. Also asked about handling continuous streams from large data platforms. The question was from an Indian interviewer, felt uneasy as they questioned if front-end roles at Google required knowledge of these topics. Seemed like the interviewer, who worked on infrastructure, viewed front-end similar to backend, with UI considered front-end in their perspective, despite me applying for a front-end role.
Lunch break.
Third Round: Behavioral questions.
Fourth Round: Pure front-end, a coding test using HTML, CSS, and JS. Given an array of numbers, had to render circles with corresponding radii. Display circles in a row, wrapping if exceeding the line. Hovering over a circle shows its area and increases the radius by 1 when clicked. Clicking a circle displays its area along with the areas of all circles to its left. Time was tight, spent time on tooltips, eventually optimized using a preSum array.
Fifth Round: A variant of a “LeetCode 汇医room2” question. Only completed this question in less than 10 minutes, but the interviewer pushed for extreme optimization, pointing out unnecessary variables. Advised against using an array, suggested using a set to avoid duplicate checks.
Rejected three days after the interview (good in behavioral questions, one strong, two weak, one bad).
Surprised by the algorithm difficulty, interviewers prompted critical thinking, mentioning that hints were equivalent to giving the answer. Even with less difficulty, optimization to the extreme was expected. Enjoyed the front-end round, engaging in discussions with the interviewer while coding, exploring multiple solutions, assessing comprehensive abilities with a large codebase, and requirements for method and code optimization.
Unlikely to interview with this company in the future, preparing for their interviews required extensive algorithm practice. The algorithm difficulty at this company does not differentiate between positions, with high-frequency questions mostly being hard or medium. Facing the interview without a strong algorithmic background was incredibly challenging. However, these experiences did not necessarily assist in interviews for front-end roles at other companies, as backend concepts, coding, and system design were also evaluated.