Staff Applied Scientist, Deep Learning Forecasting

Develop end-to-end ML/deep learning forecasting models to support planning and recommendation-related products across The Trade Desk platform. Build forecasting solutions for advertising campaigns, audience targeting, inventory management, and campaign setup using deep learning to capture large-scale signals. Design offline and online model success metrics, construct features, evaluate and monitor models, and drive iterative improvements. Collaborate with engineering on signal collection pipelines, model training, deployment, and monitoring; partner with downstream teams to understand forecasting needs and design appropriate models. Proficient in Python with Spark experience; hands-on experience building DL models at large scale using PyTorch/TensorFlow; skilled in communicating architecture decisions to cross-functional stakeholders and delivering data-powered analytics or models. Preferred background includes distributed computing with EMR/Databricks and strong data intuition. Salary is provided as a range and reflects US state-wide guidance; comprehensive benefits and equity opportunities accompany the role.