Data Science at Pinterest — The First 100 Petabytes

Jared Polivka
Kaizen Data
Published in
1 min readFeb 8, 2017
Learn about Pinterest’s data problems as they scaled from 0 to 100 Petabytes. Speaker: Mohammad Shahangian

In this video, Mohammad Shahangian walks the audience through Pinterest’s data problems as they scaled their data corpus from 0 to 100PB.

You’ll learn:

  • Critical decisions and tradeoffs that went into the original data engineering efforts at Pinterest
  • The processes and analytical methods that go into building a data driven product company like Pinterest
  • The limitations of data and the approaches Pinterest is taking to solving some of these problems

If you have any questions, post in the comments below and I’ll forward the questions on to the Pinterest team.

Meet the Speaker: Mohammad Shahangian

Mohammad Shahangian is Head of Data Science at Pinterest. Formerly, Mohammad lead Discovery Science at Pinterest where teams were responsible for making Pinterest’s billions of daily recommendations relevant. He was Pinterest’s first data scientist and initially led the development of the company’s core data infra and analytics.

--

--