Principal Data Scientist, Causal AI & Experiments

Genentech; South San Francisco (CA); Hybrid; USD 217,000 - 404,000; Full time

Job summary

Genentech, located in South San Francisco, is seeking a Principal Data Scientist to lead causal AI and experimentation efforts within the Data Science and Machine Learning team. You will design production-grade causal models, mentor colleagues, and drive measurement initiatives across CMG.

Ideal candidates have 8+ years in data science, strong Python/R, and expertise in A/B testing, MMM, uplift modeling, and discovery of business impact.

Qualifications

• 8+ years of work experience in data science, statistics, or ML engineering.
• 5+ years hands-on specialization in Causal Inference, Experimental Design, Measurement, and Optimization (e.g., A/B testing, quasi-experiments, observational causality, Market Mix Modeling, Bayesian modeling, uplift modeling).
• Expert-level knowledge of open-source frameworks for ML, statistics, and causal inference (e.g., PyTorch, TensorFlow, scikit-learn, statsmodels, CausalML, DoWhy, EconML).
• Advanced proficiency in Python and/or R, along with strong SQL skills for complex database management and data manipulation.
• Deep familiarity with open-source measurement packages (e.g., Robyn, LightweightMMM, Meridian).
• Hands-on experience scaling algorithms using big data environments (Spark, Databricks, Hadoop) and cloud platforms (AWS, GCP, or Azure).
• Exceptional verbal and written communication skills with a demonstrated ability to influence key partners, align business priorities, and explain complex mathematical methodologies to business leaders.
• Strong critical thinking abilities with rigorous attention to detail regarding experimental design, bias mitigation, and causal validity.

Responsibilities

• Drive the technical design and methodology for key Causal AI, measurement, and experimentation projects, translating complex business objectives into robust mathematical and algorithmic frameworks.
• Design, build, and deploy production-grade statistical, econometric, machine learning, and causal inference models (e.g., A/B & quasi-experimentation, Media/Marketing Mix Modeling, synthetic controls, uplift modeling, and non-parametric measurement).
• Develop scalable algorithms to quantify the true causal impact of promotional and non-promotional activities, generating actionable recommendations for business optimization.
• Partner closely with business and technical product managers, data engineers, ML engineers, MLOps, and commercial stakeholders to translate business priorities into scalable technical requirements and drive the end-to-end industrialization of Causal AI applications.
• Act as a technical authority and subject matter expert (SME) in Causal AI. Provide technical guidance, perform code and design reviews, and mentor team members to foster technical excellence and continuous learning.
• Translate complex causal modeling and statistical concepts into clear, compelling narratives and actionable guidance for both technical teams and non-technical business leaders.
• Stay at the forefront of academic and industry advancements in Causal AI, Causal ML, Bayesian modeling, and Generative AI, advocating for responsible AI practices and bringing novel approaches into our measurement ecosystem.

Skills

Python

R

SQL

PyTorch

TensorFlow

scikit-learn

statsmodels

CausalML

DoWhy

EconML

Robyn

LightweightMMM

Meridian

Spark

Databricks

Hadoop

Communication

Critical thinking

Experiment design

Education

Bachelor’s degree in a quantitative field

MS or PhD preferred

Tools

AWS

GCP

Azure

Spark

Job description

Genentech, located in South San Francisco, is seeking a Principal Data Scientist to lead causal AI and experimentation efforts within the Data Science and Machine Learning team. You will design production-grade causal models, mentor colleagues, and drive measurement initiatives across CMG.

Ideal candidates have 8+ years in data science, strong Python/R, and expertise in A/B testing, MMM, uplift modeling, and discovery of business impact.