Senior MLOps Engineer

EPAM is global leader in AI transformation engineering and integrated consulting, serving Forbes Global 2000 companies and ambitious startups. With over thirty years of expertise in custom software, product and platform engineering, we empower our clients to become AI-Native enterprises, driving measurable value from innovation and digital investments.

• 5+ years of experience in MLOps or DevOps engineering with AI/ML systems
• Hands-on experience with deployment, promotion, or rollback workflows for AI agents, including evaluation gates, version comparison, or endpoint management
• Experience in versioning and lifecycle management for LLM agents or ML models
• Experience in immutable artifact versioning at enterprise scale using AWS ECR and model registries
• Experience in CI/CD pipeline design with quality gates for AI systems
• English proficiency at B2 level or higher

Nice to have:

• Hands-on experience with AWS AgentCore Runtime versioning
• Familiarity with agent evaluation frameworks such as Langfuse, LangSmith, PromptLayer
• Experience with multi-environment agent promotion patterns
• Skills in agent version comparison and regression detection

• Define and maintain the agentic versioning model, treating agent container image, prompt configuration, tool bindings, and policy associations as an immutable unit
• Manage AWS AgentCore Runtime versioning, including automatic version creation on image push and endpoint management for DEFAULT and named endpoints
• Design and enforce the agent promotion workflow from dev to qa to prod, embedding evaluation gates guaranteeing quality before production release
• Establish agent rollback patterns that restore previous agent versions while validating functionality
• Configure AWS ECR for agent images, including ARM64 tagging, lifecycle policies, and image signing to preserve immutability
• Integrate with the Agent Registry to auto-register agent versions with metadata such as evaluation results and promotion status
• Collaborate with engineering teams to standardize versioning and lifecycle practices across the agent estate