AI/ML Platform MLOps Engineer New LATAM

Hace 2 días

central, cesar, Colombia Xebia Group Jornada completa

We are seeking a highly skilledAI/ML Platform MLOps Engineerto design, build, and operate a scalable AI/ML platform supporting the development and delivery of theLIAL product. This role focuses on enabling end-to-end machine learning lifecycle capabilities through robust platform engineering, automation, and cloud-native infrastructure.

The ideal candidate combines strong cloud engineering, ML systems expertise, and software engineering discipline to deliver production-grade AI/ML capabilities with a security-first mindset.

Key Responsibilities

  • Design, implement, and manage AI/ML platform infrastructure primarily onGoogle Cloud Platform (GCP), leveragingVertex AIservices.
  • Build and operateML pipelines using Vertex AI Pipelines, including orchestration of training, evaluation, and deployment workflows.
  • Managemodel registries, lineage tracking, and lifecycle governance.
  • Implementmonitoring and observability frameworksfor model performance, data drift, and system health.
  • Provision all infrastructure usingTerraform (GCP provider)with:
  • Remote state management

Automation & HPC Integration

  • Develop and maintainCI/CD pipelines using GitHub Actions, including:
  • Secure authentication viaWorkload Identity Federation (WIF)
  • Token exchange and secrets management
  • Author and manageSlurm-based HPC workloads, including:
  • GPU/resource allocation and scheduling
  • Hybrid workflows combining cloud and HPC environments
  • Implement automated workflows for training, validation, deployment, and retraining.
  • Develop production-grade solutions usingPython 3.10+and ML libraries such as:
  • Utilize MLOps tooling including:
  • Design, build, test, and documentREST APIsusing:
  • FastAPI or Flask
  • OpenAPI/Swagger specifications
  • Ensure API reliability, scalability, and security for integration with downstream applications.

End-to-End ML Lifecycle Ownership

  • Support the full ML lifecycle, including:
  • Feature engineering and data preparation
  • Experiment tracking and reproducibility
  • Distributed and GPU-based training workflows
  • Model evaluation, validation, and packaging
  • Deploy models toVertex AI Endpointswith scalable serving configurations.
  • Alerting and observability
  • Enableautomated retraining pipelinesbased on monitoring triggers and data changes.

Ways of Working & Collaboration

  • Operate autonomously with ownership of deliverables and platform components.
  • Apply asecurity-first mindsetacross all aspects of platform design and implementation.
  • Maintain strongplanning, documentation, and reporting discipline.
  • Collaborate effectively with:
  • Data scientists
  • Product teams (LIAL)
  • Platform and DevOps engineers
  • Continuously improve platform reliability, scalability, and developer experience.

Required Qualifications

  • Hands-on expertise with:
  • GCP (Vertex AI)and cloud-native ML workflows
  • Terraform (GCP provider)for infrastructure as code
  • Strong experience with:
  • Python (3.10+)
  • MLOps tools (MLflow, KFP SDK)
  • Experience building secure, production-grade APIs usingFastAPI or Flask
  • Solid understanding of:
  • CI/CD pipelines (GitHub Actions)
  • Identity & access management (OAuth 2.0, WIF)
  • Experience withSlurm and HPC environments, including GPU scheduling
  • Strong knowledge of:
  • ML lifecycle best practices
  • Distributed training and model deployment patterns

Preferred Qualifications

  • Familiarity with hybrid cloud/HPC architectures
  • Knowledge of data governance, compliance, and model risk management
  • Experience optimizing GPU workloads and cost efficiency on GCP
  • Exposure to scientific, analytics, or data-intensive product environments
  • Strong problem-solving and system design skills
  • Excellent communication and stakeholder engagement
  • High level of autonomy and accountability
  • Attention to detail and commitment to quality

Outcome for the Role

This role will deliver a robust, scalable AI/ML platform enabling theLIAL productto rapidly develop, deploy, and operate machine learning models in production, with high reliability, traceability, and automation across the full lifecycle.