Data Scientist - Trust & Safety

WHY DATA SCIENCE & ANALYTICS?

The Data Science & Analytics organization’s mission is to increase our speed, frequency and acumen of making decisions at scale by instilling a data-influenced approach to building products. We cover a wide area of the data spectrum including analytical data engineering, product analytics, experimentation, causal inference, statistical modeling and machine learning. Aligned and partnering with product verticals, we use this extensive tool belt to discover new opportunities and unmet use cases, influence and shape the product roadmap and prioritization, build data products and measure impact on our community of players and developers.

WHY TRUST & SAFETY?

Scaling the ability to keep our users safe amidst rapid growth is critically important at Roblox. We have several roles open across the Safety group in support of our goals to foster civility and safety across the company while allowing for maximum innovation and creative expression. Whether you’re interested in working on cutting-edge AI to detect bad actors, forecasting demand for our moderation team, identifying trends in violative user behavior, working on complex data architecture problems, building investigation tools, or understanding moderator and automated system decision accuracy, your work here will be an important driver to our mission. You will report directly to the Data Science Leader overseeing our Trust & Safety DSA team.

You Will:

• Design fundamental processes and data pipelines and curate north star metrics to track and measure the state and performance of our product and operations teams.
• Develop insights through analytics around improvements we can make in our products and processes to help build a safer platform and enable greater efficiency. 
• Work with product, engineering, and operations leaders to highlight improvements to policy and design new sets of best practices. 
• Design and run experiments to optimize safe user behavior on the platform and increase efficiency in our internal processes.
• Develop models to automate decisions in support of our goals, including automated violation detection, moderation ticket prioritization, and forecasting.

You Have:

• Bachelors or Advanced Degree in Operations Research, Computer Science, Applied Math, Physics, Engineering, Statistics, Economics or other technical fields.
• 2+ years of industry experience in data science and/or operations research 
• 2+ years of experience using big data query/processing languages such as SQL, Hive or Spark to transform/manipulate very large datasets
• 2+ years of experience in one or more scripting languages such as Python or R
• 2+ years of experience in an ever-growing company, such as the tech sector
• Demonstrated driving or building project areas from scratch as an individual contributor

You Are:

• Enthusiastic and driven. You are perpetually curious and motivated to improve your skills by learning the latest modeling techniques and data tools. You excel with freedom and autonomy and are excited by the ambiguity that comes from exploring a new problem space. 
• A strong communicator. You understand that analysis must be presented in meaningful ways and engage in spirited discussions about the findings. You have the ability to explain technical concepts to non-technical audiences and discuss appropriate tradeoffs.
• Motivated by impact. You find satisfaction in being able to point to direct contributions you’ve made to an organization. You recognize the inherent tradeoffs between further research/experimentation and productionization apply the appropriate methodology to maximize impact in a pragmatic fashion.
• An expert transforming data with SQL and a scripting language such as Python or R. You’re experienced in automating efforts and crunching massive volumes of data in distributed systems using frameworks such as Spark, Hadoop or Flink.
• Interested in optimization of large systems. You understand how to design metrics that represent the fundamental goals and constraints of the business and can design approaches to effectively trade off between countervailing factors to design more globally optimal processes.