Quora Data Science Interview
Quora has 300 Million monthly unique users.
Quora has become a platform to ask questions and connect with people who contribute unique insights and quality answers. There are 400,000 topics active on Quora right now where people can contribute. In the US alone, Quora has over 40 Million active users. Their queue currently processes 15,000 tasks per second for more than 500 types of tasks. The product relies on data and hence, data science is a very integral part of the Quora product and their business. Quora is a relatively young company (founded in 2009) but their Data Science practice and their Data Science team is very mature.
Interview Process
The interview process starts with a one hour phone screen. The phone screen has all kinds of questions from ML, A/B testing, design and statistical inference. After the phone screen, there is an onsite interview which starts with a practical assignment. The assignment starts with a dataset and expects you to find some solutions and characteristics from the dataset. There is also a coding interview to test your coding skills. The final interview is with the hiring manager which consists of high level design questions.
Important Reading
In the data and streams management layer, Quora uses Scribe for log processing (see Logging and Aggregation at Quora), Kafka for data streams and Spark for managing big data in general. In the past they were using Hadoop Map Reduce, but have now moved almost all use cases onto Spark.
- Logging and Aggregation at Quora: Logging and Aggregation
- Combatting Abuse online: Maintaining High Quality User-Generated Content through Machine Learning
- Semantic Question Matching: Semantic-Question-Matching-with-Deep-Learning
AI/Data Science Related Questions
- Write a function to return the best split for a decision tree regressor using the L2 loss.
- Print the element of a matrix in the zig-zag manner.
- How do you overcome over-fitting?
- There are X answers to a certain question on Quora. How do you create model that takes user viewing history to rank the answers? How computationally intensive is this model?
- You’re drawing from a random variable that is normally distributed X ~ N(0,1), once per day. What is the expected number of days that it takes to draw a value that’s higher that 2?
- How would you improve “Related Questions” suggestion process?
- How do you evaluate the design of the a web page?
- What are the concerns of A\B testing?
- What metrics can you come up with to measure user engagement?
- A new feature added in Quora, how will you test its effects? How will you run hypothesis testing?
Reflecting on the Questions
The data science team at Quora are great Data Science citizens of the community. They do countless presentations and share their learnings with the world. Engineering at Quora is a great place to start learning about their adventures. The interviews are extremely challenging but their team is worth it. Well rounded knowledge in the semantic and language domain can surely land you a job at one of the world’s biggest Q&A site.
Subscribe to our Acing AI newsletter, I promise not to spam and its FREE!
Thanks for reading! 😊 If you enjoyed it, test how many times can you hit 👏 in 5 seconds. It’s great cardio for your fingers AND will help other people see the story.
The sole motivation of this blog article is to learn about Quora and its technologies helping people to get into it. All data is sourced from online public sources. I aim to make this a living document, so any updates and suggested changes can always be included. Please provide relevant feedback.
