https://github.com/hectorta1989/twitter-machine-learning-engineer-interview-case-study
Twitter Machine Learning Engineer Interview | case study
https://github.com/hectorta1989/twitter-machine-learning-engineer-interview-case-study
Last synced: 8 months ago
JSON representation
Twitter Machine Learning Engineer Interview | case study
- Host: GitHub
- URL: https://github.com/hectorta1989/twitter-machine-learning-engineer-interview-case-study
- Owner: HectorTa1989
- Created: 2022-05-15T09:41:53.000Z (over 3 years ago)
- Default Branch: master
- Last Pushed: 2022-05-15T09:45:06.000Z (over 3 years ago)
- Last Synced: 2025-01-07T06:47:49.648Z (9 months ago)
- Size: 1.95 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Twitter Machine Learning Engineer Job Interview Summary
How it gets started: applied on Twitter’s website.
Job Level: SWE2
Year of Experience: 3 years experience
Degree: M.S in machine learning
Offer: Yes
TC: ~320K USD
Location: San Francisco, CA or Remote
Interview process: 4 weeks
Preparation: 3 months
Has a job: yes
Decide to join: N/A
# Twitter Machine Learning Engineer Job Interview Round 1: recruiter chat
Basic logistics questions, remote work policy, team plan for the incoming year.
# Twitter Machine Learning Engineer Job Interview Round 1: Technical Screen
1. User Active Minutes (UAM): User active minutes for a given user are defined as the count of the number of distinct minutes in which the user takes some action on Twitter.
Multiple actions in the same minute are only counted as one minute. Create a histogram of the number of users who spend X minutes on Twitter, for different values of X, given 30 days of raw logs and an interval size in minutes.
2. How will you continuously update UAM for each user by streaming?
3. How do you handle a large volume of user streaming data that can’t be fit into the memory?
# Twitter Machine Learning Engineer Virtual Onsite Interview
## Round 1: Coding and Algorithms
Given an m x n board of characters and a list of strings words, return all valid words on the board.
## Round 2: Machine learning system design
Given a trained topics model, how do you design a newsfeed machine learning system to ingest the topics model’s output as its input feature and rank all tweets for every user?
## Round 3. system design
Design a tweet scheduling service that allows user to schedule their future tweet.
## Round 4: behavioral questions
Deep dive into past machine learning projects.
Tell me about a challenging project.
How do you handle a conflict with your manager?
## Round 5: machine learning system design
Design a machine learning model to classify if a @mention is harassment, how do you collect your labels, and how do you evaluate its performance.
# Offer
Received the job offer a week after the virtual interview.