Lead Data Scientist, Video Quality

Job Description

Responsibilities
Meet Peacock, NBCUniversal's new, wildly entertaining streaming service that combines timeless shows and movies with timely news, sports and pop-culture.
We're growing our team of smart, hungry, and upbeat doers who crave the chance to build something new at the epicenter of content, tech, and culture. We need fearless leaders and pop-culture fiends who work hard and fan hard. Creative problem-solvers who just so happen to be the reigning champs at Parks & Rec trivia night. So if this sounds like you, join our flock. And we promise, we won't put your stapler in Jell-O.

Streaming video quality starts with data analysis, and Peacock is committed to delivering the best quality possible across all dimensions. In this role you will be working closely with a dedicated team to identify problems on the video side using data derived both from players and back end systems, which will include building and maintaining performance models using machine learning and other statistical techniques. You will help move the needle on the video quality of experience for millions of users, and along the way gain deep knowledge on the challenges of flawless video delivery.

Position Overview:

As part of the Direct-to-Consumer Quality and CDN Management team, the Lead Data Scientist will be responsible for creating and maintaining systems that detect anomalous behavior, find optimal configurations, and enhance user video experience on the platform.

In this role, the Lead Data Scientist will use advanced data science methodologies including collaborative filtering, deep learning, reinforcement learning, on-line modeling etc. and work closely with business owners, teammates and engineers to build a state-of-the-art real-time video streaming service.

Responsibilities include, but are not limited to:
• Lead development of analytical models using statistical, machine learning and data mining methodologies. Advise, help to resolve issues and handle non-standard cases.
• Define procedures for cleansing, discretization, imputation, selection, generalization etc. to create high quality features for the modeling process.
• Work with business stakeholders to define business requirements including KPI and acceptance criteria.
• Use big data, relational and non-relational data sources to access data at the appropriate level of granularity for the needs of specific analytical projects. Maintains up to date knowledge of the relevant data set structures and participate in defining necessary upgrades and modifications.
• Collaborate with software and data architects in building real-time and automated batch implementations of the data science solutions and integrating them into the streaming service architecture.
• Drive work on improving the codebase and machine learning lifecycle infrastructure.

Qualifications/Requirements
• Advanced (Master or PhD) degree with specialization in Statistics, Computer Science, Data Science, Economics, Mathematics, Operations Research or another quantitative field or equivalent.
• 5 years of combined experience in advanced analytics in industry or research.
• Experience in leading small teams or/and being a lead data scientist on large commercial projects.
• Deep knowledge of statistical methods and machine learning with special emphasis on the advanced algorithms like neural networks, SVM, random forests, bagging, gradient boosting machines, k-means++, deep learning or reinforcement learning. Expert level in 5+ classes of algorithms.
• Experience implementing scalable, distributed, and highly available systems using Google Cloud.
• Experience with data visualization tools and techniques.
• Understanding of algorithmic complexity of model training and testing, particularly for real-time and near real-time models.
• Proficient in at least one statistical (R, Python) and one programming (Julia, Java, Scala or similar) languages.
• Strong skills in data processing using SQL.

Desired Characteristics
• Experience with BigQuery or similar data service
• Experience with data visualizations
• Experience with reinforcement learning based systems.
• Working experience with deep learning, particularly in the areas different form computer vision.
• Experience with multi-billion record datasets and leading projects that span the disciplines of data science and data engineering
• Knowledge of enterprise-level digital analytics platforms (e.g. Adobe Analytics, Google Analytics, etc.)
• Experience building streaming data pipelines using Kafka, Spark or Flink
• Experience with large-scale video assets
• Team oriented and collaborative approach with a demonstrated aptitude and willingness to learn new methods and tools
• Pride and ownership in your work and confident representation of your team to other parts of NBCUniversal

Jobcode: Reference SBJ-dy8bnp-3-143-244-83-42 in your application.