Responsibilities
- Use statistical and ML based techniques to reduce selection bias, build representative samples, and analyze randomized studies in the ad-tech space.
- Work hand-in-hand with our engineering team to design and implement models that measure ad performance across large, disparate datasets.
- Perform R&D work by prototyping new statistical and data modeling frameworks, translating prototypes into SQL, pandas-based and Pyspark workflows, and then drive those prototypes into production.
- Extend a well-structured codebase that leverages Python class abstractions and modular pipeline design; write clean, testable measurement logic.
- Troubleshoot or maintain internal models by understanding the key ingredients and underlying assumptions of the models.
- Build and maintain models that deliver in-depth ad campaign measurement inside privacy-preserving clean room environments.
- Translate statistical models into configuration-driven, production-ready PySpark workflows parameterized by different configurations.
- Partner with software and data engineers to design, monitor, and improve the cloud infrastructure that powers end-to-end measurement pipelines across multiple cloud environments.
- Drive new product development in the retail media/brand measurement space - from prototyping new model methodologies to shipping configurable measurement workflows that operations teams can run at scale across multiple customers..
- Engage with internal and external clients to understand their measurement needs and translate those insights into pipeline improvements and new product offerings.
- Design, develop, and maintain statistical and ML-based measurement models that run at scale in privacy-preserving clean room environments.
- Translate measurement methodology into robust, configuration-driven production pipelines using PySpark and SQL, ensuring correctness through rigorous data quality checks and invariant validation.
- Prototype new statistical and data modeling frameworks - from exploratory research through to production-ready workflows - iterating rapidly while maintaining code quality and reproducibility.
- Apply causal inference and bias correction techniques to address selection bias, build representative national samples, and analyze randomized and observational studies.
- Collaborate closely with data and software engineers to integrate measurement models into cloud-based data infrastructure, and troubleshoot automated pipeline failures end-to-end.
- Partner with product, solutions, and operations teams to translate client measurement requirements into configurable pipeline parameters and new product capabilities.
- Validate and audit model outputs across large, disparate datasets - identifying anomalies, diagnosing root causes, and ensuring results are statistically sound and defensible to clients.
- Document methodology, model assumptions, and operational runbooks; lead training sessions to enable internal teams to independently operate and interpret measurement workflows.
- Stay current with industry best practices in ad-tech measurement, privacy-preserving computation, and causal inference — bringing new ideas back to improve the platform.
Basic qualifications
- Masters with 5-8 plus years experience or PhD 2+ years experience in a data science related field (eg. Statistics, Mathematics, Computer Science, Engineering, Economics, or a related discipline).
- Strong proficiency in Python (pandas, PySpark) and SQL for working with large, complex datasets across distributed environments.
- Solid statistical foundation - regression, classification, time-series, sampling, and selection bias correction; hands-on experience designing and analyzing experiments (A/B, holdout, geo-tests) and applying causal inference methods.
- Experience with modern data and analytics frameworks (Spark, Jupyter, Airflow) and cloud environments (AWS or GCP). Understand object-oriented programming, modular pipeline design, version control (git), and comfort with command-line interfaces.
- Experience building, prototyping, and productionizing statistical or ML-based models - translating research into robust, maintainable production workflows.
- Familiarity with digital advertising measurement concepts (attribution, reach, ROAS, conversion modeling) or strong curiosity to learn ad-tech measurement, able to translate complex measurement insights into clear, actionable recommendations for technical and non-technical stakeholders - including product teams, and internal/external customers.
- Comfortable collaborating across technical stakeholders - data engineers, solution, architects, and product managers - with strong written and verbal communication skills.
- Strong problem-solving mindset: first-principles thinker, intellectually curious, rigorous, and adaptable to evolving tools and industry best practices.
- Proficiency using AI coding assistants and LLM-based tools (such as Claude, Cursor, or GitHub Copilot) to accelerate research, prototyping, and code development; ability to write effective prompts and critically evaluate AI-generated outputs. Familiar with agentic AI workflows - orchestrating multi-step AI pipelines, using tool-calling frameworks, or building prompt-driven automation to support data science and operational tasks.
Preferred qualifications
- Excellent communication skills and the ability to work with internal stakeholders to translate product requests into released features.
- Familiarity with privacy preserving data environments - data platforms, clean rooms, or data collaboration environments.
- Experience in adtech, martech, or digital marketing analytics, including topics such as addressable media, identity resolution, attribution, or media measurement.
- Familiarity with retail / CPG data - store banners, product hierarchies, transaction-level datasets, and experience with privacy-preserving data environments (clean rooms, k-anonymity, aggregation thresholds) or large disparate datasets spanning impressions, transactions, and audiences.
Benefits
- People: Work with talented, collaborative, and friendly people who love what they do.
- Fun: We host in-person and virtual events such as game nights, happy hours, camping trips, and sports leagues.
- Work/Life Harmony: Flexible paid time off, paid holidays, options for working from home, and paid parental leave.
- Comprehensive Benefits Package: LiveRamp offers a comprehensive benefits package designed to help you be your best self in your personal and professional lives. Our benefits package offers medical, dental, vision, life and disability, an employee assistance program, voluntary benefits as well as perks programs for your healthy lifestyle, career growth and more.
- Savings: Our 401K matching plan—1:1 match up to 6% of salary—helps you plan ahead. Also Employee Stock Purchase Plan - 15% discount off purchase price of LiveRamp stock (U.S. LiveRampers)
Tags & focus areas
Used for matching and alerts on DevFound Fulltime Remote Data Science Ai