In the five days from July 24th to 28th 2017, I interviewed at LinkedIn, Salesforce Einstein, Google, Airbnb, and Facebook, and got all five job offers.
It was a great experience, and I feel fortunate that my efforts paid off, so I decided to write something about it. I will discuss how I prepared, review the interview process, and share my impressions about the five companies.
How it started
I had been at Groupon for almost three years. It’s my first job, and I have been working with an amazing team and on awesome projects. We’ve been building cool stuff, making impact within the company, publishing papers and all that. But I felt my learning rate was being annealed (read: slowing down) yet my mind was craving more. Also as a software engineer in Chicago, there are so many great companies that all attract me in the Bay Area.
Life is short, and professional life shorter still. After talking with my wife and gaining her full support, I decided to take actions and make my first ever career change.
Preparation
Although I’m interested in machine learning positions, the positions at the five companies are slightly different in the title and the interviewing process. Three are machine learning engineer (LinkedIn, Google, Facebook), one is data engineer (Salesforce), and one is software engineer in general (Airbnb). Therefore I needed to prepare for three different areas: coding, machine learning, and system design.
Since I also have a full time job, it took me 2–3 months in total to prepare. Here is how I prepared for the three areas.
Coding
While I agree that coding interviews might not be the best way to assess all your skills as a developer, there is arguably no better way to tell if you are a good engineer in a short period of time. IMO it is the necessary evil to get you that job.
I mainly used Leetcode and Geeksforgeeks for practicing, but Hackerrank and Lintcode are also good places. I spent several weeks going over common data structures and algorithms, then focused on areas I wasn’t too familiar with, and finally did some frequently seen problems. Due to my time constraints I usually did two problems per day.
Here are some thoughts:
This area is more closely related to the actual working experience. Many questions can be asked during system design interviews, including but not limited to system architecture, object oriented design, database schema design, distributed system design, scalability, etc.
There are many resources online that can help you with the preparation. For the most part I read articles on system design interviews, architectures of large-scale systems, and case studies.
Here are some resources that I found really helpful:
Many blogs are also a great source of knowledge, such as Hacker Noon and engineering blogs of some companies, as well as the official documentation of open source projects.
The most important thing is to keep your curiosity and modesty. Be a sponge that absorbs everything it is submerged into.
Machine learning
Machine learning interviews can be divided into two aspects, theory and product design.
Unless you are have experience in machine learning research or did really well in your ML course, it helps to read some textbooks. Classical ones such as the Elements of Statistical Learning and Pattern Recognition and Machine Learning are great choices, and if you are interested in specific areas you can read more on those.
Make sure you understand basic concepts such as bias-variance trade-off, overfitting, gradient descent, L1/L2 regularization，Bayes Theorem，bagging/boosting，collaborative filtering，dimension reduction, etc. Familiarize yourself with common formulas such as Bayes Theorem and the derivation of popular models such as logistic regression and SVM. Try to implement simple models such as decision trees and K-means clustering. If you put some models on your resume, make sure you understand it thoroughly and can comment on its pros and cons.
by Xiaohan Zeng, Medium | Read more:
It was a great experience, and I feel fortunate that my efforts paid off, so I decided to write something about it. I will discuss how I prepared, review the interview process, and share my impressions about the five companies.
How it started
I had been at Groupon for almost three years. It’s my first job, and I have been working with an amazing team and on awesome projects. We’ve been building cool stuff, making impact within the company, publishing papers and all that. But I felt my learning rate was being annealed (read: slowing down) yet my mind was craving more. Also as a software engineer in Chicago, there are so many great companies that all attract me in the Bay Area.
Life is short, and professional life shorter still. After talking with my wife and gaining her full support, I decided to take actions and make my first ever career change.
Preparation
Although I’m interested in machine learning positions, the positions at the five companies are slightly different in the title and the interviewing process. Three are machine learning engineer (LinkedIn, Google, Facebook), one is data engineer (Salesforce), and one is software engineer in general (Airbnb). Therefore I needed to prepare for three different areas: coding, machine learning, and system design.
Since I also have a full time job, it took me 2–3 months in total to prepare. Here is how I prepared for the three areas.
Coding
While I agree that coding interviews might not be the best way to assess all your skills as a developer, there is arguably no better way to tell if you are a good engineer in a short period of time. IMO it is the necessary evil to get you that job.
I mainly used Leetcode and Geeksforgeeks for practicing, but Hackerrank and Lintcode are also good places. I spent several weeks going over common data structures and algorithms, then focused on areas I wasn’t too familiar with, and finally did some frequently seen problems. Due to my time constraints I usually did two problems per day.
Here are some thoughts:
- Practice, a lot. There is no way around it.
- But rather than doing all 600 problems on Leetcode, cover all types and spend time understanding each problem thoroughly. I did around 70 problems in total and felt that was enough for me. My thought is that if 70 problems isn’t helpful then you may not be doing it right and 700 won’t be helpful either.
- Go for the hardest ones. After those the rest all become easy.
- If stuck on one problem for over two hours, check out the solutions. More time might not be worth it.
- After solving one problem, check out the solutions. I was often surprised by how smart and elegant some solutions are, especially the Python one-liners.
- Use a language that you are most familiar with and that is common enough to be easily explained to your interviewer.
This area is more closely related to the actual working experience. Many questions can be asked during system design interviews, including but not limited to system architecture, object oriented design, database schema design, distributed system design, scalability, etc.
There are many resources online that can help you with the preparation. For the most part I read articles on system design interviews, architectures of large-scale systems, and case studies.
Here are some resources that I found really helpful:
Although system design interviews can cover a lot of topics, there are some general guidelines for how to approach the problem:
- Understand the requirements first, then lay out the high-level design, and finally drill down to the implementation details. Don’t jump to the details right away without figuring out what the requirements are.
- There are no perfect system designs. Make the right trade-off for what is needed.
Many blogs are also a great source of knowledge, such as Hacker Noon and engineering blogs of some companies, as well as the official documentation of open source projects.
The most important thing is to keep your curiosity and modesty. Be a sponge that absorbs everything it is submerged into.
Machine learning
Machine learning interviews can be divided into two aspects, theory and product design.
Unless you are have experience in machine learning research or did really well in your ML course, it helps to read some textbooks. Classical ones such as the Elements of Statistical Learning and Pattern Recognition and Machine Learning are great choices, and if you are interested in specific areas you can read more on those.
Make sure you understand basic concepts such as bias-variance trade-off, overfitting, gradient descent, L1/L2 regularization，Bayes Theorem，bagging/boosting，collaborative filtering，dimension reduction, etc. Familiarize yourself with common formulas such as Bayes Theorem and the derivation of popular models such as logistic regression and SVM. Try to implement simple models such as decision trees and K-means clustering. If you put some models on your resume, make sure you understand it thoroughly and can comment on its pros and cons.
by Xiaohan Zeng, Medium | Read more:
[ed. If you're considering a degree in computer programming/engineering and feel drawn to Silicon Valley, good luck! (... then you have to find a place to live.)]
