Preparing for Machine Learning Engineer role Vs Software Engineer role in India

Anonymous User

14892

Google Microsoft Salesforce Data Science Career interview experience India Machine Learning Engineer Amazon

I recently got offers from top product based companies FAANG and an unicorn startup for machine learning engineer role in India. I would like to talk about my preparation strategies for machine learning engineer role and the patterns I came across interviewing at top companies. A little addition to your preparation might fetch you machine learning role.

I have around 2 years of experience in machine learning at a medium sized product based company in Bangalore. I've Bachelor's degree in computer science from tier-3 college (2018 passout). I got SDE role from few and SDE-2 role from a company. I was planning to go for master's to USA this year but due to covid I dropped my plans and decided to join a new company here. I started my preparation in January and started applying for jobs in May and ended up getting few offers(some rejects too) by July end (Interview process is slower these days).

Machine Learing roles in companies:

Research based:
* Ex: Research Scientist, Applied scientist, Research Engineer
* Typical requirements: PhD in related field, 3-5 years of experience, degree from tier-1 college
* Salary : Slightly higher (30 LPA + )
Application based:
* Ex: SDE -1 (Machine learning), SDE -2 (Machine Learning)
* Typical requirements: 1-2 years of experience (Usually they expect less experience in Software Engineering however they expect you know basic machine learning)
* Salary : Slightly lower (18 LPA + )

I can't say about research based role as I didn't apply for them however I can certainly say about the application based roles. The process is slightly different than typical SDE interviews.

Screening round :

Usually 2 rounds (1 DSA + 1 ML)

Culture fit and behavioural round:

As ML/AI teams usually follow different standards in different companies you should expect one round

Usually 3 rounds technical data strcutures and system design

1 system design interview (mostly HLD for example: Design google search. Here you need to talk about ML algorithms and also about systems, caching etc)
2 data structures and algorithms rounds:
* 2 questions in each round (detailed and also expect to write tests for your code)
* You should surely expect one dbms question in either of the rounds.

Around 6-7 rounds in total (Expect 2 rounds more than usual SDE roles). I'm taking about most companies not all.

Typical difficulty of questions:

DSA questions are slightly easier than typical non-ML SDE roles (Leetcode medium)--Leetcode hard questions are very rare. However this is compensated with math in ML specific rounds.
System design (Expect HLD)
Expect Math, probablity, model evaluation and loss functions based questions
Database management is a must topic.

It took me around 5 months of preparation to crack the interviews. I've solved aroud 270 questions on leecode (easy+medium).

It is not difficult to crack ML related roles as I was able to crack these interviews in 6 months. Feel free to comment if you need any information, ML preparation resources or questions, I will be happy to answer.

Edit 1: Added ML specific interview preparation resources

Comments (12)