MLS Engineer

Must-Have Qualifications

• 69 years of applied ML/data science experience, with a track record of models shipped to production and measured via online experiments.
• Statistics: Advanced grounding in statistics and probability theory — hypothesis testing, Bayesian inference, causal inference, and experimental design (power analysis, confounding, novelty effects) — applied rigorously to ambiguous, real-world problems.
• Algorithms: Deep understanding of machine learning algorithms and their mathematical foundations (linear algebra, optimization theory, computational complexity); able to select, adapt, or design algorithms rather than default to out-of-the-box implementations.
• Deep Learning & Optimization: Solid grounding in neural network architectures (sequence models, transformers, embeddings) and optimization methods (SGD variants, regularization, hyperparameter tuning); able to reason about convergence and generalization trade-offs.
• Hands-on proficiency in Python (pandas, scikit-learn, PyTorch or TensorFlow) and SQL; comfortable working in Jupyter/Databricks notebooks against large datasets.
• Experience with distributed data processing (Apache Spark) at a scale of tens/hundreds of millions of rows.
• Demonstrated ability to design and interpret A/B tests; familiarity with causal inference methods (propensity scoring, uplift modeling, diff-in-diff) is a strong plus for pricing/marketing-adjacent work.
• Research Literacy: Comfortable reading and critically evaluating ML/statistics literature, and translating relevant papers into working methods rather than relying solely on established techniques.
• Software Craftsmanship: Writes clean, version-controlled, testable research code (Git workflows, code review, reproducible experiments) suitable for handoff to ML Engineers.
• Event-Driven Systems: Development and deployment experience with event-driven architectures — microservices and streaming pipelines (e.g., Kafka, event-driven model serving) — for productionizing and scaling models in real-time systems.
• Ability to communicate modeling trade-offs to non-technical stakeholders.

Nice-to-Have

• Advanced degree (MS/PhD) in Statistics, Computer Science, Applied Math, or a related quantitative field, with research experience in algorithm design or statistical learning theory.
• Experience with recommendation/ranking systems (learning-to-rank, embeddings, two-tower models) or demand forecasting.
• Familiarity with feature stores and model registries (e.g., MLflow, Feast-style feature stores).
• Exposure to AWS SageMaker or similar managed ML platforms.
• Background in travel, e-commerce, or marketplace pricing/search problems.
• Publications, Kaggle competitions, or open-source ML contributions.