About the Role
TripleLift is seeking an experienced Staff Data Scientist with a focus on optimization strategies for our real-time ad marketplace. In this role, you will lead research and experiments to understand bidding behaviors, improve the performance of bidding algorithms, and optimize outcomes for Publishers, DSPs, and TripleLift. You will build proof-of-concept models, deploy ML models into production, create reusable features and data structures, and collaborate with cross-functional teams to advance Data Science initiatives.
Responsibilities
- Lead research and experiments to improve the performance of our bidding algorithm and maximize performance for our customers.
- Conceptualize new bidding and pricing optimization strategies based on cutting-edge research, build proof-of-concept ML models, and drive them to successful outcomes in full-scale production.
- Drive analytics projects in partnership with Product, Engineering, and cross-functional teams to support and influence product strategies.
- Monitor and measure statistical modeling performance, and build dashboards and alerts to ensure models are functioning effectively.
- Serve as the technical Subject Matter Expert (SME) on all aspects related to bidding strategies and marketplace dynamics.
- Build reusable modules and data structures, and provide guidance and feedback to team members on their work, taking into account their skills, backgrounds, and working styles.
Qualifications
- Master's degree or higher in a related quantitative field (e.g., Mathematics, Computer Science, Engineering, Economics, or Operations Research).
- At least five years of work experience in a relevant role building pricing optimization strategies for web-scale systems.
- Documented history of conceptualizing new ideas based on cutting-edge research and productionalizing them to drive successful outcomes.
- High level of proficiency across a majority of tools used in our tech stack which includes Python, Spark, DataBricks, ONNX, MySQL, Snowflake, Airflow, Docker, and Amazon Web Services.
- Deep knowledge of ML libraries like scikit-learn to quickly analyze data and prototype models that can be used in high volume distributed systems.
- Familiarity with monitoring and measuring statistical modeling performance with ability to quickly build dashboards and alerts, using tools like Prometheus, Grafana, Looker, etc., to make sure models are functioning effectively.
- Experience deploying and maintaining web-scale, production-grade, Data Science systems.
- Excellent technical communication skills.
- Strong analytical and problem-solving skills.
- Committed to a process of continuous learning and spreading subject matter expertise widely through talks, blog posts, or written documentation.
Technologies Used by the Data Science Team
- Languages: Python
- Frameworks: Spark, DataBricks, ONNX, Docker, Airflow
- Databases: MySQL, Snowflake, S3/Parquet
- Cloud Platform: Amazon Web Services