Senior AI / MLOps Engineer – e-Commerce Search & Information Retrieval

5+ years of experience in software engineering. 2+ years focused on deploying ML/AI systems at scale. Strong coding skills in Python. Hands-on expertise with containerization (Docker). Hands-on expertise with orchestration (Kubernetes/EKS/GKE/AKS). Hands-on expertise with cloud platforms (AWS, GCP, or Azure). Proven record of building CI/CD pipelines and automated testing for data or ML workloads. Deep understanding of REST/gRPC APIs. Deep understanding of message queues (Kafka, Kinesis, Pub/Sub). Deep understanding of stream/batch data processing frameworks (Spark, Flink, Beam). Experience implementing monitoring, alerting, and logging for mission-critical services. Familiarity with common ML lifecycle tools (MLflow, Kubeflow, SageMaker, Vertex AI, Feature Store, etc.). Working knowledge of ML concepts such as feature engineering, model evaluation, A/B testing, and drift detection.

Convert prototypes into production-ready Python and Go microservices or libraries. Design reproducible build pipelines and manage artifacts. Orchestrate real-time and batch inference workloads on Kubernetes or managed services. Implement blue-green/canary rollouts and model versioning strategies. Build and maintain CI/CD pipelines with automated testing. Automate feature store updates, model retraining, and batch jobs. Define and monitor SLIs/SLOs for model performance and cost. Integrate logging, tracing, and metrics with alerting and on-call practices. Collaborate on scalable data pipelines for various data types. Implement real-time and offline feature extraction and validation. Profile models and services for performance and cost optimization. Right-size clusters and workloads for performance and cost balance. Embed security, privacy, and responsible-AI checks in pipelines. Manage secrets, IAM roles, and data-access controls. Ensure auditability and reproducibility through documentation. Partner with Data Scientists, Product Owners, and SREs. Coach junior engineers on MLOps best practices.