Senior Machine Learning Engineer, Advertiser Growth

Unity Unity · Enterprise · New York, NY · AI & Machine Learning

Senior Machine Learning Engineer role focused on building GenAI systems for creative asset generation, designing budget pacing algorithms, developing marketplace experimentation infrastructure, and ensuring high-scale billing reliability within Unity's ad-tech ecosystem. The role requires strong backend systems experience and leadership in a high-growth environment.

What you'd actually do

  1. Design and optimize sophisticated pacing controllers (PID, probabilistic forecasting) to smooth advertiser spend across diverse time zones and traffic spikes, ensuring optimal delivery and marketplace stability.
  2. Lead the backend integration of Generative AI models (Diffusion, LLMs) to automate the creation of high-performing image and video assets tailored to specific campaign goals and formats.
  3. Build and scale the infrastructure for high-velocity experimentation, including A/B testing, switchback tests, and long-term holdouts to measure the impact of marketplace changes on advertiser ROI and platform health.
  4. Architect and maintain high-throughput billing pipelines that process billions of events with 100% accuracy, bridging the gap between real-time ad delivery and mission-critical financial reconciliation.
  5. Analyze complex financial and marketplace datasets to refine the trade-off between spend velocity and advertiser performance, using experimentation results to tune pacing and billing logic.

Skills

Required

  • 4+ years of software engineering experience
  • 1+ year working on ads delivery systems
  • building and operating large-scale, low-latency backend systems
  • building or maintaining budget control systems, feedback loops, or spend-prediction algorithms
  • building or scaling experimentation platforms
  • working on "mission-critical" pipelines (like billing, payments, or clearinghouses)
  • building the backend workflows required to serve, scale, and store Generative AI models for creative asset generation
  • real-time stream processing (Kafka, Flink, or Spark)

Nice to have

  • Experience embracing AI as a strategic advantage in engineering

What the JD emphasized

  • mission-critical pipelines
  • zero-fault tolerance
  • 100% accuracy

Other signals

  • GenAI systems for creatives
  • budget pacing algorithms
  • experimentation infrastructure
  • high-scale billing reliability