Senior Applied Machine Learning Scientist, Content Demand Modeling (NLP)

Key Skills

Apache Hadoop, C Programming, C++, Data science techniques, Java Programming, Machine learning techniques, MapReduce, Python Programming, PyTorch, R Programming, Scala Programming, SQL, TensorFlow

Job Description

Netflix is revolutionizing the entertainment industry with world-class technology. We are both a content distributor and a producer for original and premium shows. We serve millions of subscribers in more than 190 countries around the world. We produce hundreds of new series, movies, documentaries, stand-up specials, and content across various other categories each year. Because of our global footprint, we are able to elevate new types of creators, tell a diverse set of stories and inspire a global audience.

The Content Demand Modeling team, within Data Science and Engineering, plays a central role in informing and supporting content decisions. Arming our creative teams with data-driven insights improves both the efficiency and accuracy of decision-making and helps Netflix produce more hits and create more member joy at better economics. Our shared mission is to scale decision-making, opportunity detection, and discovery for creative exploration.

We would like to bring on board a machine learning scientist with strong natural language processing and natural language understanding experience. In this role, you will unlock the potential of unstructured text-based assets (e.g. Hollywood screenplays, scripts, outlines, synopses) to deepen our understanding of how our members engage with a show and to support planning related to the production of our content. You will use world-class machine learning techniques on real-world data to directly impact the evolution of our content catalog.

Questions You Will Help Answer

• What was happening in a show at times when viewership dramatically changed (e.g., dropoffs)? What aspects correlate with sustained engagement from our members?
• Are there narrative or structure elements related to the story that support a stronger sense of connection from our members?
• What places, people and themes are referenced in a screenplay/script - and how can this information be used to support more efficient production planning, including scheduling and resource allocation (e.g. for visual effects)?

Responsibilities

• Develop and execute on the overall roadmap for model-driven content understanding, identifying milestones & checkpoints that support iteration, learning & progress
• Develop proof of concept solutions to draw early feedback from creative executives and content and studio strategy partners
• When appropriate, partnering with Product teams & Engineering to get models deployed and operational in end-to-end solutions used by the Netflix Studio
• Develop foundational machine learning models that enable parsing and modeling of text assets (e.g. named entity recognition to support a host of script-based ML applications)
• Extend techniques that traditionally work well for English language text assets to support other languages in which Netflix creates and produces content

About You

• You are excited about using machine learning to unlock insights about how our members engage with our content or how our content is produced. You are energized by the opportunity to use ML to understand content.
• You have strong experience with Natural Language Processing / Understanding and/or Computational Linguistics.
• You thrive in ambiguity and have strong prioritization skills. You combine a strong sense of business impact with great judgment around how to make iterative progress.
• You have an MSc or PhD with 6+ years of Applied ML research and development. You have deployed large-scale, innovative ML solutions that have successfully delivered business value from conception to production.
• You excel in at least one major programming language (e.g., Python) and ML/DL framework (e.g., PyTorch, TensorFlow, Keras).
• You have strong communication skills to collaborate with cross functional partners to drive research. You have a track record of thought leadership (e.g. deploying great ML products, peer-reviewed publications, talks at conferences, etc.)
• You have experience with SQL (any variant), big-data tech (e.g.,Spark, Hadoop, Hive) to connect data with ML pipelines.