April 30, 2015

Facebook Shares Data Science Career Insights with CDS

Some of the members from Facebook’s Data Science Team visited the Center of Data Science in March. The Team included Ali Cramer, who is responsible for Data Science Team’s recruitment, as well as two practitioners in Data Science, Hyan Park and Anurag Jain. They came there to discuss the roles of Data Engineers and Data Scientists at Facebook, in addition to career opportunities for both full time positions and interns.

Data Science at Facebook

Hyan Park gave us an overview of her role at Facebook and the responsibilities of a data scientist at any company. She noted that data scientists are supposed to support the decisions of the company with data and analysis. This involves asking thoughtful questions and then answering those questions with supporting data. She suggested that data scientists should be asking:

  1. How would changing of the application’s metric affect the usability and other aspects?
  2. How do I efficiently track the effects of the decisions being made?
  3. What is the impact of the new feature additions in these metrics?
  4. Where is the overall ecosystem headed?

Ms. Park also noted some of the data science projects she has worked on including:

  • Shifting of a metric’s location on the mobile application: this sometimes resulted in positive changes like an increased usability of application. But sometimes negatively affected revenue as it decreased ad clicks.
  • Improving the quality of suggestions to people: An example of this would be related to managing the “follows” on Instagram. Analyzing the current trends of followings to answer various questions helps Facebook improve the suggestions’ quality.

Data Engineering at Facebook

Anurag Jain, a data engineer at Facebook, shared some of his experiences and responsibilities. The role of a data engineer is to focus across products to build a stable data pipeline, as well as a platform to retrieve and analyze data. Since Facebook is a global application and is expected to show almost real time response, there is a strong need to update the data very fast – to all instances across the world!

A data engineer’s role is very much like a software engineer’s with the main difference being that data engineers focus solely on data stream and the related platforms.

Innovation at Facebook

In addition to the data scientist roles and responsibilities, the Facebook team provided information on its “innovative culture”.  Facebook proactively encourages innovation through a division called Creative Labs, which is a product development team that works on several innovative products at once.  They also organize and run Hackathons  to facilitate any new ideas from employees.

The talk generated many questions including:

How the company maintains data privacy from the employees working on the data?

How much data that they used to perform various data science analysis and the techniques they used? and

How would they start analyzing if there was a sudden problem in any of the expected results?

Overall it was an interesting session and excited everyone in the room about the potential opportunities at Facebook – particularly those related to large, real world datasets.

Story by Rishabh Jain, master’s degree candidate and writer for the Center for Data Science.