JOBSEARCHER

MLOps Engineer

Senior MLOps Engineer (GCP) — ML Platform Framework We are seeking a Senior MLOps Engineer to architect and build our next-generation, end-to-end ML Platform-as-a-Product. In this role, you will design the unified framework thatabstracts infrastructure complexities, enabling ML Engineers and Data Scientists to seamlessly write, train, deploy, and monitor models in production. Key ResponsibilitiesML Lifecycle StageCore Focus & Deliverables Authoring & Local DevStandardize notebook environments, local CLI tools, and SDKs to make model writing highly repeatable and containerized from day one. Training & Orchestration Build automated, scalable training pipelines utilizing Vertex AI Pipelines, Kubeflow, and Databricks, integrating unified data access layers. Productionalization Construct robust CI/CD/CT (Continuous Training) pipelines that package models into auto-scaling microservices on Google Kubernetes Engine (GKE) or Vertex AI Endpoints. Operations & Drift Implement end-to-end observability frameworks to track system performance, monitor data/concept drift, and trigger automated alert-and-retrain loops. Platform Governance Manage the enterprise Feature Store, artifact registries, and unified metadata tracking (lineage) for compliance and governance. Technical Requirements & Stack • Platform Architecture: Proven experience building a unified "ML Platform-as-a- Service" inner-source framework for internal engineering consumption. • GCP Ecosystem: Deep expertise in Vertex AI (Pipelines, Feature Store, Model Registry, Endpoints), GKE, BigQuery, and Google Cloud Storage. • Infrastructure as Code (IaC): Strong hands-on experience using Terraform to provision secure, compliant, and cost-optimized ML infrastructure. • Maturity Frameworks: Experience setting up end-to-end automation covering ML GitOps, automated testing, and zero-downtime canary deployments. Key Takeaway Your objective is simple: Reduce time-to-market for ML models from months to minutes by delivering a friction-free, standardized, and self-service engineering framework