Machine Learning Engineer

About The Role

The role owns the design, training, and production deployment of machine learning models that power core product features at scale - spanning NLP, ranking, and predictive systems processing millions of daily inferences.

You will work in a small, senior team where infrastructure decisions, model quality, and shipping velocity all matter equally, and your work directly shapes how the platform learns from user behavior.

Key Responsibilities

• Design, train, and evaluate ML models for production use cases including ranking, classification, and forecasting, with clear ownership of offline and online metrics
• Build and maintain training and feature pipelines in Python and PySpark, ensuring consistency between training and serving environments
• Deploy models end-to-end using Kubernetes, TorchServe, or managed cloud services (SageMaker, Vertex AI), including canary rollouts and rollback strategies
• Instrument model monitoring for data drift, latency, and prediction quality; build automated alerting and retraining triggers
• Optimize inference performance - quantization, batching, model distillation - to meet strict latency and cost budgets
• Collaborate with data engineering, product, and applied science teams to translate ambiguous problems into well-scoped ML solutions
• Contribute to code reviews, design docs, and shared ML platform tooling used across the organization

What We Are Looking For

• 3–6 years of experience in machine learning engineering, applied ML, or a closely related role, with multiple models shipped to production
• Strong Python engineering skills; comfortable writing production-grade, tested code beyond notebooks
• Deep hands-on experience with PyTorch or TensorFlow, and modern ML tooling (MLflow, W&B, Kubeflow, Airflow)
• Production experience with cloud ML infrastructure on AWS, GCP, or Azure, including containerized deployment (Docker, Kubernetes)
• Solid ML fundamentals: evaluation methodology, feature engineering, regularization, and the failure modes of models in production
• MS or PhD in Computer Science, Statistics, or a related quantitative field, or equivalent industry experience
• Bonus: experience with LLM fine-tuning, distributed training (DeepSpeed, FSDP), real-time streaming systems (Kafka, Flink), or Rust/C++ inference optimization