Staff Applied Scientist

The Trade Desk The Trade Desk · Media · London, United Kingdom · Data Science

Staff Applied Scientist role focused on building and scaling AI/ML systems for digital advertising, including control systems, recommendation engines, and time series models. Requires strong ML, statistics, and experimental design skills, with experience in end-to-end project ownership from research to production.

What you'd actually do

  1. Build and maintain control systems that dynamically allocate ad budgets across sell-side partners based on quality signals and performance targets.
  2. Design and analyze experiments using causal inference methods to measure the impact of inventory decisions on advertiser KPIs, and develop the metrics that make those trade-offs visible and actionable.
  3. Build the recommendation systems and curation frameworks that help buyers find supply that reaches their target audiences at scale and delivers on their campaign outcomes.
  4. Develop time series models to anticipate supply volume shifts and help buyers and partners plan accordingly.
  5. Partner with engineering, product, and business teams to translate research into scalable, production-ready systems that improve marketplace value for buyers and sellers.

Skills

Required

  • machine learning
  • statistics
  • experimental design
  • causal inference
  • metric development
  • recommendation systems
  • ranking models
  • large‑scale data processing
  • Spark
  • EMR
  • Databricks
  • clean code
  • version control
  • code review

Nice to have

  • programmatic advertising
  • real-time auctions
  • supply-side systems
  • control theory
  • constrained optimization
  • budget allocation systems
  • agentic AI workflows
  • LLMs for automation
  • decision support
  • data-driven product features

What the JD emphasized

  • owning projects end-to-end
  • productionization at scale
  • large‑scale data processing

Other signals

  • build and maintain control systems
  • design and analyze experiments using causal inference methods
  • build the recommendation systems and curation frameworks
  • develop time series models
  • translate research into scalable, production-ready systems