Sr Applied Scientist

Uber Uber · Consumer · Amsterdam, Netherlands · Data Science

Senior Applied Scientist at Uber to build and deploy ML/AI solutions in production, taking ideas from concept to real-world systems. The role involves end-to-end work from problem definition to production integration, focusing on classification, prediction, anomaly detection, and risk scoring. It also includes improving the reliability and robustness of AI systems, including LLM-based applications, and applying model adaptation techniques like fine-tuning.

What you'd actually do

  1. Build and deploy ML/AI models in production across a range of problem areas (e.g., classification, prediction, anomaly detection, risk scoring)
  2. Work end-to-end: problem definition → modeling → evaluation → production integration
  3. Collaborate with engineers to integrate models into scalable, reliable systems
  4. Design experiments and define metrics to measure performance and impact
  5. Continuously improve models based on data, feedback, and real-world usage

Skills

Required

  • Ph.D., MS, or Bachelor’s degree in a quantitative field (CS, Statistics, Engineering, etc.)
  • 2+ years (PhD/MS) or 4+ years (BS) of industry experience in Applied ML / AI
  • Strong foundation in machine learning and statistics
  • Experience building and deploying ML systems in production
  • Proficiency in Python and working with large datasets
  • Experience with experimentation, evaluation, and data analysis

Nice to have

  • Experience with LLMs, including: RAG systems, prompt design, and evaluation
  • Experience with model adaptation techniques, such as: fine-tuning, parameter-efficient tuning (e.g., LoRA, adapters)
  • Familiarity with reinforcement learning or feedback-driven optimization approaches
  • Experience working with large-scale data systems (e.g., Spark, Hive, Presto)
  • Familiarity with reliability or safety considerations in AI systems
  • Experience in domains such as security, fraud, or risk

What the JD emphasized

  • ML/AI models in production
  • scalable, reliable systems
  • metrics
  • reliability and robustness of AI systems
  • fine-tuning
  • parameter-efficient tuning
  • feedback-driven optimization
  • LLMs
  • RAG systems
  • prompt design
  • evaluation
  • model adaptation techniques

Other signals

  • deploy ML/AI models in production
  • work end-to-end
  • integrate models into scalable, reliable systems
  • improve the reliability and robustness of AI systems