Senior Machine Learning Engineer, Vector Bidding Science

Unity Unity · Enterprise · Mountain View, CA · AI & Machine Learning

Senior Machine Learning Engineers to join the Vector Bidding Science team, focusing on defining technical vision and architecting next-generation scalable real-time bidding systems powered by AI, marketplace intelligence, and advanced optimization frameworks. The role involves developing state-of-the-art bidding and pacing algorithms to maximize advertiser returns by harnessing massive datasets and rich signals.

What you'd actually do

  1. Design, implement, and optimize core bidding algorithms and auction mechanisms
  2. Architect and scale bid landscape forecasting capabilities
  3. Analyze large-scale marketplace dynamics to uncover deep insights and deliver algorithmic improvements
  4. Drive offline evaluations and online A/B experiments to validate model performance and deliver measurable business impact
  5. Collaborate cross-functionally with product, infrastructure, and engineering teams

Skills

Required

  • state-of-the-art machine learning
  • reinforcement learning
  • control theory
  • complex, real-world bidding or pricing problems
  • software engineering skills in Python
  • deep learning frameworks, preferably PyTorch
  • metric design
  • online experimentation frameworks (A/B testing)
  • large-scale data analysis
  • lead projects end-to-end
  • deliver measurable business impact
  • ambiguous technical landscape

Nice to have

  • large datasets
  • distributed computing frameworks (e.g., Spark, Ray, BigQuery, Flink)
  • real-time ad systems
  • DSP (Demand-Side Platform) bidding logic
  • survival analysis for market price estimation
  • game-theoretic modeling
  • AI tools (such as Claude Code, GitHub Copilot, and Cursor)

What the JD emphasized

  • complex, real-world bidding or pricing problems
  • ambiguous technical landscape

Other signals

  • real-time bidding systems
  • cutting-edge AI
  • advanced optimization frameworks
  • massive datasets
  • state-of-the-art bidding and pacing algorithms