Senior Machine Learning Engineer

Roku•Austin, TX

13h•Hybrid

About The Position

Roku is the #1 TV streaming platform in the U.S., Canada, and Mexico, aiming to power every television in the world. The Advertising Performance group focuses on performance for Advertisers, Publishers, and Roku, utilizing Machine Learning, Reinforcement Learning, AI, Control and Optimization Systems, and Auction Dynamics to solve complex problems. This involves real-time multi-objective optimization with distributed systems at large scale and low latencies. At the core is their Machine Learning, Experimentation, and Inference Platform, which is continuously evolved. In this role, you will apply various methodologies to solve challenging problems in Advertising, including conversion modeling, calibration, dynamic creative generation and optimization, forecasting, timeseries modeling, yield and margin optimization, and experimentation for A/B and multivariate testing. You will also contribute to building a state-of-the-art machine learning platform.

Requirements

BS or higher in CS, ECE or a related field.
5+ years of experience in building out Machine Learning platforms or applying Deep Learning methodologies.
Ability to communicate and collaborate cross functionally.
Solid grasp of core statistical techniques.
Deep experience in SOTA Deep Learning discriminative and generative models.

Nice To Haves

MS or higher in CS, ECE or a related field.
Experience in the Advertising domain.
Contributions to open-source ML projects.

Responsibilities

Building SOTA Deep learning discriminative models and generative models to generate image and video ads geared towards optimizing performance.
Building and evolving a SOTA Machine Learning Platform from feature generation to realtime inferencing that can optimize and deploy complex models at large scale and low latencies.
Being forward-thinking about advancements in ML related areas.