Data Scientist Lead Consultant - Arity

at  Arity

At Allstate, great things happen when our people work together to protect families and their belongings from life’s uncertainties. And for more than 90 years our innovative drive has kept us a step ahead of our customers’ evolving needs. From advocating for seat belts, air bags and graduated driving laws, to being an industry leader in pricing sophistication, telematics, and, more recently, device and identity protection.

JOB DESCRIPTION

Arity
Founded by The Allstate Corporation in 2016, Arity is a data and analytics company focused on improving transportation. We collect and analyze enormous amounts of data, using predictive analytics to build solutions with a single goal in mind: to make transportation smarter, safer, and more useful for everyone. At the heart of that mission are the people that work here—the dreamers, doers and difference-makers that call this place home. As part of that team, your work will showcase both your intelligence and your creativity as you tackle real-world problems and put your talents towards transforming transportation. That’s because at Arity, we believe work and life shouldn’t be at odds with one another. After all, we know that your unique qualities give you a unique perspective. We don’t just want you to see yourself here. We want you to be yourself here. Arity is committed to supporting an inclusive and diverse environment where you can thrive and learn from others.
The Team
Advertising data science team empowers the intelligence and efficiency of Arity Marketing Platform, which enables marketers to find their best customers wherever they are in their journey. Our goal is to empower marketers to reach prospects and display relevant messaging based on how, when, and where they drive. The team is responsible for optimizing the in-house marketing platform for advertisers and publishers, as well as our revenue, throughout all funnels. To achieve this, not only do we need to look for inefficiencies on the platform including individual campaigns, but also how to grow and scale the platform to gain more revenue. This means we need to find a cost-effective way to rapidly deploy, test and iterate algorithms on our platform. This team is also fully integrated within a cross-functional scrum structure including product owners, software engineers, ad operation team, and other talents to collaborate with to achieve the same business goals

As an advertising data scientist at Arity, you will lead the development of machine learning algorithms on both 1st party driving data like GPS and driving events, as well as, ad platform data including impressions, clicks, conversions. You are expected to make key technical decisions about how to implement machine learning models including both traditional and deep learning models on large volumes of data. You have chances to influence the business for the entire cycle of the optimization solution including data collection, processing, modeling, A/B testing on the ad platform. You will personally prototype these solutions and work with product owners, data/software engineers, and other partners for productionization and KPI measurement. Some example projects include:

• Click-through rate (CTR)/conversion prediction
• Win rate model
• Dynamic bidding strategy
• Pacing control and budget management
• Frequency capping model
• Platform simulation

These models help us understand how effective our ad platform is and provide opportunities for improvement and growth. You will also help shape and grow our culture we have worked hard to establish – promoting recognition of good work, continuous learning, winning together, and having fun along the way.

QUALIFICATIONS

Successful candidates typically have:

• Master’s or PhD degree in a machine learning/AI related field such as engineering, statistics, computer science, physics, or related discipline
• 5+ years of industry experience in data science, data analytics, and machine learning in advertising domain
• Deep experience in digital advertising including systems, measurement sciences, and principled incrementality approaches and passion for incentive challenges
• Experience with ad auctions (RTB platform) such as dynamic bidding strategy, ad ranking, and experimentation on the advertising platform
• Advanced knowledge in predictive models such as parameterized methods, ensemble algorithms, deep neural network, and reinforcement learning algorithms such as multi-armed bandit algorithms, Q-learning, deep reinforcement learning
• Demonstrated experience using Python and Spark for big data query/processing and engineering skills for productionizing the solution
• Experience with scientific computing libraries Scikit-learn, TensorFlow, PyTorch, and Spark ML-lib
• Over 5 years’ experience with developing end-to-end machine learning solutions/algorithms including model development, deployment, monitoring, and life-cycle management on the advertising platform
• Ability to translate product requirement into well-defined analytical problems and produce feasible solutions
• Ability to provide written and oral interpretation of highly specialized terms and data, and ability to present this data to stakeholders with different levels of expertise

Optional:

• Experience with control theory and the application to ad platform optimization
• Experience with cloud data warehouse solutions such as BigQuery or Redshift
• Experience with deploying ML models using AI platforms such as Vertex AI and Sagemaker
• Experience with geospatial data is preferred, such as US census data, weather data, parcel data, POI, etc

SKILLS

Algorithms, Big Data, Data Analytics, Data Science, Deep Learning, Digital Advertising, Machine Learning, Machine Learning Algorithms, Predictive Modeling

Your day-to-day looks like:

• Be a thought partner in the area of experimentation for Ads Platform, and autonomously identify and pursue research with significant business impact on KPIs
• Analyzing large amount of data sets using distributed computing frameworks
• Building advanced algorithm and machine learning models using a variety of libraries/tools and cutting-edge techniques
• Identifying opportunities for new machine learning solutions, exploring new data sources for enrichment, collecting appropriate labels for learning, establishing actionable metrics, and creating reusable model validations and risk mitigation
• Communicating results to key stakeholders in a clear and compelling manner
• Establishing and following data science best practices including peer review, code review, documentation, coding standards, and ensuring reproducibility and compliance
• Working with product and engineering partners for model handoff and productionization