Build operational tooling for model lifecycle management across environments.
Design and support enterprise feature stores to improve reuse and governance.
Integrate ML solutions with data warehouses, data lakes, and operational sources.
Develop scalable data pipelines for training and inference.
Optimize data availability, quality, lineage, and observability.
Implement monitoring for model performance, data quality, and drift.
Establish alerting, observability, and incident management processes.
Develop strategies for retraining and continuous improvement.
Build and support low-latency scoring and prediction platforms.
Develop event-driven ML architectures using streaming technologies.
Implement MLOps governance standards and controls with security and compliance alignment.
Support auditability, explainability, lineage tracking, and documentation.
Collaborate with cross-functional teams and mentor junior engineers.
Drive adoption of modern ML engineering methodologies.
Experience Requirements
5+ years in Machine Learning Engineering, MLOps, Data Engineering, or Software Engineering.
3+ years building and supporting production ML systems.
End-to-end ML lifecycle management experience.
Cloud experience with AWS, Azure, or GCP.
Experience with batch and real-time ML solutions and large-scale data.
ML and MLOps: lifecycle management, deployment and orchestration, experiment tracking and versioning, monitoring and observability, feature store and model registry, automated retraining, responsible AI and governance.
Cloud platforms: AWS SageMaker, Azure ML, or Google Vertex AI.
Programming: Python and SQL required; Java, Scala, or R preferred.
Data platforms: Snowflake, Databricks, IBM DB2 or equivalent, data lakes and cloud storage.
Data transformation: dbt, Spark, Pandas, PySpark.
DevOps and infrastructure: Git-based workflows, CI/CD tools, Terraform or IaC, Linux, shell scripting.
Containers and orchestration: Docker, Kubernetes, OpenShift preferred.
Streaming and real-time: Apache Kafka, event-driven architectures, real-time inference pipelines.
Monitoring and observability: Prometheus, Grafana, ELK, cloud-native solutions.
Preferred: enterprise feature stores, model registry and governance frameworks, real-time decisioning platforms, credit risk or fraud or customer analytics or recommender systems, MLFlow or Kubeflow or Airflow or Argo or Feast or SageMaker Pipelines, distributed systems, AI/ML regulatory frameworks, relevant cloud certifications.
