Machine Learning Engineer (Foundation Models & Personalization)
Eight Sleep8 months ago
Responsibilities
- Build and deploy ML models for personalization, prediction, behavior understanding, readiness forecasting, event detection, and individualized recommendations.
- Apply foundation-model capabilities to product workflows, including LLMs, tools, retrieval-augmented generation, multimodal modeling, policy learning, and MCP-style integrations.
- Develop user behavior models connecting longitudinal sleep, environment, and routine signals to actionable interventions.
- Design offline evaluation strategies using metrics, slice-based analysis, calibration, reliability, and fairness measures.
- Partner with Product to run online experiments and measure product impact.
- Build scalable training and inference pipelines and productionize models.
- Implement model monitoring, drift detection, alerting, and continuous improvement loops.
- Collaborate with Product, Mobile, Backend, and Clinical partners to scope and ship high-impact features.
Requirements
- At least 2 years of experience building production machine learning systems, ideally for consumer-facing products.
- Strong fundamentals in supervised learning, sequence and time-series modeling, and modern deep learning.
- Hands-on experience with large-scale model training and evaluation using PyTorch, TensorFlow, or JAX.
- Strong Python engineering practices.
- Experience with personalization systems, including ranking, recommendations, segmentation, lifecycle modeling, propensity modeling, behavior modeling, and causal or experiment-aware thinking.
- Fluency with SQL, distributed compute such as Spark or Ray, and cloud storage or compute.
- Ability to translate ambiguous product goals into measurable outcomes and iterate with stakeholders.
- Preferred experience applying LLMs or foundation models to product features, including tool use, retrieval, structured outputs, guardrails, or evaluations.
- Preferred experience with multimodal data, sensor signals, context, health data, or biometrics data.
- Preferred experience with privacy-preserving approaches such as on-device learning, federated learning, differential privacy, or data minimization.
- Preferred experience designing experimentation frameworks or causal inference approaches for personalization.
Benefits
- Equity participation for every full-time team member, including periodic performance-based equity refreshes.
- A Pod product provided to every employee.
- Immediate responsibility and opportunities for accelerated career growth.
- Collaboration with experienced talent in a fast-paced innovation-focused environment.
Tech Stack
Categories
About Eight Sleep
Eight Sleep builds the Pod, a smart mattress system that controls temperature, elevation, and wake-up and provides sleep analytics for consumers and athletes. The company sells hardware directly online with a companion app and optional membership for personalized insights and automations. Founded in 2014 and headquartered in New York, it is privately held and sells in the U.S., Canada, the U.K., Europe, Australia, the UAE, Mexico, Switzerland, and Saudi Arabia.