Machine Learning Engineer

Role overview A remote machine learning role for an engineer who cares about what happens after the notebook. The position owns ML problems end to end, from framing the question through data work, training, deployment, and ongoing monitoring, with a focus on data quality and honest evaluation as first-class concerns.

Responsibilities - Frame business problems as ML problems, then build, evaluate, and ship the models that solve them. - Own the full lifecycle, including data pipelines, training, deployment, and monitoring, not just the modeling step. - Build evaluation harnesses that accurately reveal whether a model is actually working in production. - Watch for drift and degradation, and be ready to retrain or roll back before clients feel the impact. - Work directly with clients and product to ensure models solve the real problem rather than a proxy metric.

Requirements - 4+ years building and deploying machine learning systems that have run in production. - Strong Python skills and fluency with the practical ML stack, such as PyTorch or TensorFlow, scikit-learn, and pandas. - Solid grounding in data modeling, feature engineering, and honest evaluation, with awareness of leaky-metric pitfalls. - Clear communication of trade-offs and a preference for shipping a simpler model that works over a clever one that does not.

Nice to have - Experience with MLOps tooling and serving models reliably at scale. - Comfort with cloud infrastructure and data pipelines end to end. - A paper, a competition placing, or a production model you are genuinely proud of.

Benefits and work setup - Real ownership of ML problems from framing through production, with no ticket queue. - Fully remote, async-friendly setup with real overlap hours, plus hardware and compute budget and a yearly learning stipend. - Meaningful equity so engineers share in what they build.