Sep 26, 2018
Welcome back to Tech Forward, listeners! On this week’s episode, I spoke with Julie Pitt, director of Machine Learning Infrastructure at Netflix. Her team’s goal is to scale Data Science while increasing innovation. Prior to that role, she built the streaming infrastructure behind the "play" button while Netflix was transitioning from domestic DVD-by-mail service to international streaming service. Julie also co-founded Order of Magnitude Labs, with a mission to build AI capable of doing things like exploration, communication, and accomplishing long-range goals — in other words, tasks that humans find easy, but today’s machines find hard. Today, we’ll talk about Netflix’s unique culture and how Julie supports the other female engineers on her team.
Julie’s work with machine learning at Netflix ties back to solving two problems: offering members the best choices, and ensuring they have as much control as possible over their personal experience. This requires answering a lot of questions, from where to find content creators, to predicting where that content will be popular, and even down to launch dates. Luckily, many of those questions can be answered through data science and machine learning. Software engineers have already experienced the progression from rigid to continuous deployment schedules, and witnessed firsthand the power of collaboration. That power, however, is not common in data science — which is where Julie’s team comes in. They’re building a workshop for data scientists that allows a higher degree of collaboration, more efficiency, and higher productivity. Netflix’s data scientists have already used these workshops to solve a variety of problems, from determining the quality of the video streamed to the viewer, to optimizing a schedule for launching multiple titles in the same genre.
One unique aspect of Netflix’s culture, according to Julie, is that “We’re building a default platform, but not a mandated platform.” This is a direct result of two main tenets of the Netflix culture: freedom and responsibility. While engineers have the freedom to make choices about the tech they use to solve problems, they’re also expected to be aware of how those choices might affect other teams. Julie also champions the concept of leadership through context, not control. “As a leader, if someone on my team makes a poor decision, my first instinct is to determine what context they lacked. We can increase relevant context that enables better decisions.”
When it comes to diversity initiatives at Netflix, Julie has personally focused on meetups for underrepresented groups, as well as rethinking the interview process from start to finish. She gives candidates a choice as to how they prefer to be interviewed: some people feel more comfortable with a technical whiteboard question, while others prefer to work independently on a project and discuss it afterwards. Once people are in the door, having a network is key — “I only ever had one job where I didn’t already know someone. I cannot under-emphasize the importance of having a professional network.” She also has some words of advice for anyone struggling with impostor syndrome: “Fundamentally, in tech, we operate in an environment of uncertainty because we’re doing things that have never been done before. Everyone else is making it up as they go along, too. It’s okay, and you’re not alone.”
Julie, thank you so much for coming onto the show and sharing your stories and insights this week. I’d also like to thank all of you out there listening, reviewing, and sharing the show. See you next week!
Connect with us