Senior ML Infrastructure Engineer (DevOps)

About the role

We are seeking a Senior ML Infrastructure Engineer who thrives on running production systems at scale. You will partner closely with applied scientists to turn their models into reliable, real‑time services for millions of customers.

Key responsibilities

• Maintain real‑time inference services, respond to alerts, debug incidents, perform rollbacks and write post‑mortems as part of a 24x7 on‑call rotation.
• Design, build, and improve AWS infrastructure using IaC, CI/CD pipelines, Kubernetes, Docker, autoscaling, monitoring and cost‑control mechanisms.
• Manage the GPU fleet – driver and CUDA upgrades, node provisioning, capacity planning and troubleshooting.
• Automate user access, SSH credentials, service accounts and developer environments for the science team.
• Run real‑time data pipelines, stream ingestion into an online feature store, and serve low‑latency feature endpoints.
• Orchestrate batch workloads with Apache Airflow or Databricks Workflows.
• Collaborate with scientists to productionise prototypes and keep them healthy.
• Write and review production‑grade Python code, including AI‑generated snippets.

Required profile

• Proven 24x7 on‑call experience with high‑load online services and live incident resolution.
• Strong Linux systems expertise and hands‑on experience with Kubernetes, Docker and Infrastructure as Code.
• Solid Python, Git and CI/CD background.
• Experience with workflow orchestration tools such as Airflow or Databricks Workflows.
• Familiarity with Gen AI development lifecycle (SDLC) and English proficiency at B2 level or higher.

Required skills

• Python
• Git
• CI/CD pipelines
• AWS
• Infrastructure as Code (e.g., Terraform, CloudFormation)
• Kubernetes
• Docker
• Linux
• NVIDIA driver & CUDA
• Apache Airflow
• Databricks Workflows

What we offer

• Remote work from anywhere in Georgia with optional office use in Tbilisi, Batumi or Kutaisi.
• Opportunity to work abroad up to two months per year.
• Relocation possibilities across 55+ EPAM offices worldwide.
• Leadership development, career advising, soft‑skill and well‑being programs.
• Certification support for GCP, Azure and AWS.
• Unlimited access to EPAM’s internal learning platform and free English classes.
• Participation in the Employee Stock Purchase Plan.