Senior / Staff Data Scientist (IC4/IC5) - Yield Optimization

Responsibilities

• Lead the research and modeling behind the Yield Optimization portfolio: Demand Shaping, Floor Optimization and Bid Optimization.
• Turn business problems in programmatic auctions into well-posed modeling problems, with clear objectives, constraints and success metrics such as revenue and margin per bid request.
• Build and validate models for bidding, reserve prices and traffic shaping, using auction theory, causal inference, forecasting and online learning methods such as bandits.
• Know the bid-stream data in depth: find biases, leakage and data-quality problems, and design features that hold up in production.
• Identify and champion new methods to maintain Verve's competitive edge in yield optimization.
• Plan and execute large-scale initiatives with significant impact, managing multiple interdependencies and cross-team collaborations.
• Work with ML engineers to take models to production: set the latency and feature requirements, review the serving design, and own model quality after launch.
• Act as a technical leader and mentor to other data scientists and engineers across distributed teams.
• Refine team problem-solving tools, methodologies, and technical documentation to uplift the overall team competency.
• Partner closely with Product Management, ML Engineering and Backend Engineering to align long-term technical roadmaps with business strategy.
• Stay hands-on: write production-quality model code, run the statistical analysis behind major decisions, and design and read A/B tests and offline evaluations.

Required Skills

• 5+ years in data science or applied ML, with models you built running in production and a measurable business impact.
• Extensive AdTech industry expertise, with deep knowledge of programmatic auctions, supply/demand matching, and real-time bidding optimization.
• Strong foundations in statistics, probability and optimization, with applied experience in at least one of auction theory, causal inference or experimental design.
• Master-level proficiency in Python and SQL.
• Deep expertise in ML methods and frameworks (Scikit-learn, gradient boosting, PyTorch or TensorFlow), including model selection, validation and offline evaluation.
• Hands-on analysis of large datasets with BigQuery and PySpark, and model training on a cloud platform such as Vertex AI.
• Working knowledge of real-time model serving, enough to design models that meet latency and feature constraints.
• Exceptional communication skills with the ability to influence technical and non-technical stakeholders across the organization.
• Ability to navigate and align multiple teams around long-term technical visions and modeling standards.
• Advanced degree (Master's or PhD preferred) in a quantitative field (Applied Mathematics, Statistics, Computer Science, or Physics).