Senior Data Scientist - Programmatic Algorithms

About impact.com

impact.com is a leading commerce partnership marketing platform that revolutionizes business growth by enabling comprehensive partnership management across the customer journey. The company connects brands with affiliates, influencers, content publishers, brand ambassadors, and customer advocates to drive trusted, performance-driven growth through strong relationships. Its flagship products Performance (affiliate), Creator (influencer), and Advocate (customer referral) combine all partner types into a singular platform. Currently, over 5,000 global brands including Walmart, Uber, Shopify, Lenovo, L’Oréal, and Fanatics leverage impact.com to support more than 350,000 partnerships delivering measurable business outcomes.

Position Overview

We are hiring a Senior Data Scientist to become an embedded expert within our Programmatic Experience Group.This critical role involves architecting and deploying machine learning models that optimize pricing, inventory allocation, and yield at large scale. Sitting at the crossroads of data science, platform engineering, and marketplace economics, you will take full ownership of your projects—building data pipelines, engineering features, and launching real-time inference systems that enable quick business decisions. Your contributions are vital to balancing advertiser success with publisher revenue in Impact's programmatic marketplace, making this a high-impact technical position.

Key Responsibilities

• Develop ML models for auction pricing, bid shading, floor price determination, and inventory yield optimization.
• Maintain real-time pricing algorithms that harmonize immediate revenue goals with long-term stakeholder health.
• Implement adaptive feedback loops to allow pricing models to respond to evolving market dynamics and inventory changes.
• Analyze revenue effects of pricing improvements and clearly present trade-offs to stakeholders between yield, fill rates, and ROI.
• Manage ML strategies for inventory allocation including routing, pacing, and matching supply with demand across various segments and deal types.
• Forecast inventory supply, demand curves, and price clearing to guide proactive allocation decisions.
• Identify and resolve inefficiencies such as unsold inventory or uneven deal matching in publisher base.
• Design and manage data infrastructure supporting programmatic models, including event pipelines, feature stores, and real-time feature serving.
• Engineer impactful features from large-scale auction logs, bid streams, user indicators, contextual data, and historic performance metrics.
• Build durable data pipelines emphasizing reliability, monitoring, version control, and efficient data reprocessing.
• Deploy models in real-time inference environments ensuring low-latency, robustness, and high throughput for auction decision-making.
• Establish monitoring frameworks for model efficacy, data drift, system performance and set alerts and retraining triggers.
• Collaborate with MLOps and platform teams to meet service level objectives under intense auction traffic.
• Manage full model lifecycle including training, validation, deployment, and A/B testing.
• Create and run rigorous controlled experiments to measure model impact on yield, fill rates, advertiser KPIs, and publisher income.
• Develop evaluation frameworks that account for live auction dynamics and delayed feedback signals.
• Communicate experimental results with actionable insights to product and business stakeholders.
• Research and introduce self-learning and adaptive algorithms such as contextual bandits and reinforcement learning to enhance the programmatic stack.
• Design automated feedback systems linking auction results to model updates, minimizing manual adjustments.
• Keep abreast of the latest advances in programmatic ML, auction theory, and online optimization to refine model strategies.
• Act as main ML technical collaborator for related product and engineering teams, transforming requirements into feasible modeling approaches.
• Work alongside data science colleagues on common infrastructure, modeling practices, and feature sharing.
• Document models, methods, and experimental outcomes professionally for transparent knowledge sharing and reproducibility.

Candidate Qualifications

• Minimum 5 years in data science, ML engineering, or quantitative fields, including at least 2 years applying ML in programmatic advertising, ad tech, or related domains involving real-time bidding, pricing or auction systems.
• Deep knowledge of programmatic auction mechanisms such as RTB, header bidding, floor pricing, deal types, and how machine learning can optimize supply-demand dynamics.
• Strong track record in production ML engineering: independently taking models from development to real-time deployment with monitoring and retraining pipelines.
• Experience designing high-volume, low-latency data architectures including pipelines, feature stores, and training data management.
• Proficiency in Python and SQL and familiarity with ML frameworks like scikit-learn, XGBoost, LightGBM, PyTorch, or TensorFlow; plus usage of big data tools like Spark and Kafka.
• Expertise in real-time model serving technologies (REST, gRPC, streaming inference) and comprehensive production ML workflows such as drift detection, versioning, and experiment management.
• Capability to handle and model programmatic datasets with large cardinality and high event volumes.
• Strong understanding of causal inference and experimental design for online systems with delayed and noisy feedback conditions.
• Excellent communication skills for explaining complex concepts and collaborating in multidisciplinary teams.
• Academic background: Bachelor's degree in quantitative disciplines (e.g., Computer Science, Statistics, Mathematics, Engineering, Economics). Advanced degrees preferred.

Preferred Additional Skills

• Prior practical experience with supply-side platforms, demand-side platforms, or exchange yield optimization including floor pricing and bid landscape modeling.
• Familiarity with auction theory especially first-price/second-price auctions and revenue optimization concepts.
• Experience applying reinforcement learning, multi-armed bandits, or contextual bandits to real-time decision-making challenges.
• Knowledge of adaptive online learning algorithms designed for non-stationary environments.
• Understanding of privacy-centric ML methods such as differential privacy, federated learning, or cookieless attribution techniques.
• Experience with Google Cloud Platform tools like BigQuery, Vertex AI, Dataflow, Pub/Sub and/or Databricks for large-scale data processing.
• Background in supply forecasting, inventory control, or capacity planning within programmatic or marketplace contexts.
• Familiarity with impact.com's affiliate and partnership ecosystem or experience at the convergence of performance marketing and programmatic delivery.

Distinguishing Attributes

• An economic mindset toward marketplaces appreciating incentives, equilibria, and systemic effects of modeling decisions.
• End-to-end mastery of ML workflows from feature engineering to model deployment and system-level troubleshooting.
• Innovative feedback loop design enhancing model self-improvement and minimizing manual interventions.
• Analytical rigor in imperfect real-world environments managing delayed and noisy signals with practical trade-offs.
• Results-oriented delivery balancing research depth with implementable MVP solutions enabling fast iteration.
• Strong collaborative engagement building trust and alignment with engineering and product teams through transparent and committed execution.

Compensation & Benefits

The salary range for this role is between 165,000 and 185,000 USD annually, along with a performance-dependent 5% variable bonus and eligibility for Restricted Stock Unit (RSU) grants subject to board approval and a standard 3-year vesting.

impact.com offers an extensive benefits package designed to promote employee wellness and work-life balance, including comprehensive health, vision, and dental insurance; life and disability coverages; flexible paid time off policies; mental health supports with therapy and coaching sessions; gym reimbursement; stock ownership opportunities; continuous learning through Coursera and PXA courses; generous parental leave; technology stipends for home office setup; and internet expense allowances. Note that benefits may vary based on location and applicable local laws.

Equal Opportunity

impact.com is an equal opportunity employer, providing equitable treatment and employment opportunities regardless of racial, ethnic, or ancestral background, or color.