Lead, Data Science Operations

Echo Global Logistics•Chicago, IL

2d•$129,352 - $188,077•Remote

About The Position

The Data Science Operations Lead is a senior individual-contributor role situated at the intersection of Data Science, Engineering, and IT Architecture. This position focuses on the operational aspects of the model lifecycle, including deployment, monitoring, scaling, and maintenance. The role is crucial for ensuring the reliability, observability, and governance of Echo's growing portfolio of production models, allowing Data Scientists to concentrate on developing new capabilities. The Lead will act as the team's expert for transitioning models from research and development to production services, serve as the primary contact for production issues, and liaise with the Architecture team on deployment and reliability matters.

Requirements

Hands-on experience operating ML or software systems in production: an MLOps, DevOps, SRE, platform, or data science background with demonstrated production ownership.
Strong working knowledge of CI/CD pipelines, deployment automation, and a major cloud platform (AWS, Azure, or GCP).
Demonstrated expertise in error handling, fault tolerance, and designing systems that fail gracefully (retries, fallbacks, alerting, monitoring/observability).
Proficiency in Python (R a plus), and a working understanding of how ML models are packaged, served, monitored, and retrained.
Comfort serving as first point of contact for production issues, including an on-call / off-hours expectation.
A teaching disposition, with the ability to translate complex infrastructure into clear guidance for colleagues who are not infrastructure specialists.

Nice To Haves

Experience standing up monitoring and observability for a portfolio of production models or services (e.g., drift detection, performance tracking, alerting).
Familiarity with containerization (Docker) and orchestration (Kubernetes), infrastructure-as-code, and model-serving frameworks.
Familiarity with MLOps tooling such as MLflow, Airflow, or Kubeflow, or managed equivalents (e.g., SageMaker, Vertex AI), and with data/model versioning.
Experience working across an engineering/architecture boundary as a liaison or embedded operations partner.
Pragmatic use of AI tooling to accelerate operations and code-quality work, paired with sound judgment about when human reasoning is required.

Responsibilities

Model deployment partnership: Serve as Data Science's primary counterpart to the Architecture / Platform Engineering team on model deployment, owning day-to-day collaboration, hand-offs, and coordination to bridge the gap between trained models and production services (APIs, web tools).
Production reliability and incident response: Act as the first point of contact for production issues (outages, errors, degraded endpoints) across all deployed models and endpoints, including on-call / off-hours availability.
Resilient, error-aware systems: Bring rigor to error handling and fault tolerance, designing and enforcing practices to prevent errors and ensure graceful degradation or failure of models and endpoints with sensible fallbacks, retries, alerting, and recovery paths.
Monitoring and observability: Establish and maintain the necessary monitoring and observability for managing a portfolio of production models as an enterprise capability, tracking model health, endpoint performance, latency, logging, and prediction quality.
Deployment expertise and team enablement: Develop a deep understanding of the evolving deployment system, acting as the team's guide, helping Data Scientists move from experiment to production quickly and safely, and driving templating, documentation, and automation.
Governance and quality: Own versioning, reproducibility, and operational governance for models in production, partnering with Architecture on standards and controls to ensure a trustworthy model and algorithm footprint.