Automate the model lifecycle end to end through CI/CD and GitOps: packaging, promotion across environments, progressive rollout, and rollback.
Build and operate production model-serving infrastructure for LLMs and predictive models, including inference optimization, autoscaling, and low-latency serving across multiple model formats and runtimes.
Architect and manage GPU infrastructure — scheduling, autoscaling, resource isolation, and utilization efficiency for training and inference workloads.
Instrument the platform and the models on it — structured logging, telemetry, drift detection, and cost tracking — and build the triggers and pipelines that close the loop into retraining and revalidation.
Extend model, dataset, and artifact registries, metadata systems, and versioning so every deployed model has a traceable, auditable history.
Build the developer-facing surface of the platform: APIs, SDK components, templates, documentation, and runbooks that make it self-service, backed by well-tested code and active participation in design and code review.
Uphold company mission and values through accountability, innovation, integrity, quality, and teamwork.
Maintain regular and reliable attendance.
Act with an inclusion mindset and model these behaviors for the organization.
Minimum Qualifications
Bachelor's degree in Computer Science, Engineering, AI/ML, or a related field; or equivalent practical experience.
3+ years building and operating production software, including significant work on ML or AI infrastructure.
Strong Python, including software engineering fundamentals — automated testing, code review, and designing code others will read and extend.
Experience with the ML model lifecycle: pipelines that carry a model from training through evaluation, deployment, monitoring, and retraining.
Kubernetes experience — deploying, scaling, and debugging containerized workloads on a major cloud provider (AWS preferred).
Experience with CI/CD, GitOps-based delivery, and infrastructure automation.
Preferred Qualifications
ML workflow orchestration and experiment tracking (for example Kubeflow, Argo Workflows, Airflow, MLflow, or Weights & Biases).
GPU infrastructure at scale: scheduling, multi-tenancy and resource isolation, autoscaling, and inference optimization such as quantization, batching, or graph compilation (for example ONNX Runtime or TensorRT).
Feature stores, data versioning, or dataset lineage tooling.
Progressive delivery for models — canary, shadow, or A/B deployment and automated rollback.
Model monitoring and drift detection, including automated retraining triggers.
Model governance and reproducibility practices: lineage tracking, audit trails, and approval workflows.
Kubernetes-native serverless and event-driven autoscaling (for example Knative or KEDA).
Experience building APIs or shared libraries consumed by other engineering teams.
Production experience in an additional systems language such as Go or Java.
Fine-tuning foundation models, or building distributed training pipelines.
Experience delivering software in a regulated environment (HIPAA, CLIA, GxP, SOC 2).
We are an equal employment opportunity employer. All qualified applicants will receive consideration for employment without regard to age, color, creed, disability, gender identity, national origin, protected veteran status, race, religion, sex, sexual orientation, and any other status protected by applicable local, state, or federal law. Applicable portions of the Company’s affirmative action program are available to any applicant or employee for inspection upon request.
- The base pay for this position is
- $61,300.00 – $122,700.00
- In specific locations, the pay range may vary from the range posted.
JOB FAMILY
Product Development
DIVISION
ONCO Cancer Diagnostics
LOCATION
United States > Madison : 5505 Endeavor Ln
ADDITIONAL LOCATIONS
WORK SHIFT
Standard
TRAVEL
Yes, 10 % of the Time
MEDICAL SURVEILLANCE
No
SIGNIFICANT WORK ACTIVITIES
Continuous sitting for prolonged periods (more than 2 consecutive hours in an 8 hour day), Keyboard use (greater or equal to 50% of the workday)
Abbott is an Equal Opportunity Employer of Minorities/Women/Individuals with Disabilities/Protected Veterans.
EEO is the Law link - English: http://webstorage.abbott.com/common/External/EEO_English.pdf
EEO is the Law link - Espanol: http://webstorage.abbott.com/common/External/EEO_Spanish.pdf
