Machine Learning Operations (MLOps) Engineer

About this role

Gallatin seeks an MLOps Engineer to build production infrastructure for AI/ML systems supporting U.S. national security logistics. You'll own the full pipeline from model training to deployed capability across cloud and on-premises environments, including air-gapped and restricted networks.

What you'll do

• Design and operate model training, fine-tuning, and inference infrastructure on AWS (SageMaker, EKS) and on-premises GPU hardware
• Build LLM serving stacks (vLLM, quantization, batching) with local-first deployment for degraded-mode resilience
• Develop CI/CD pipelines with versioned datasets, model registry, promotion gates, and reliable rollback mechanisms
• Create evaluation harnesses with regression testing, LLM-as-judge systems, and adversarial datasets; instrument production for drift and silent failures
• Manage data ingestion, embedding pipelines, feature freshness, and human-in-the-loop confidence routing and feedback systems
• Deploy and operate ML systems in IL5/IL6 environments; support ATO/continuous authorization with security control evidence (NIST, CMMC, FIPS 140-3, RMF/eMASS)

What they're looking for

• MLOps and ML platform engineering (5+ years production experience)
• Python, Kubernetes, containerization, and infrastructure as code
• AWS ML services (SageMaker, EKS) and cloud ML infrastructure
• GPU scheduling, memory optimization, and cost management
• IL5/IL6 and air-gapped/restricted-network deployment operations
• LLM inference serving and optimization (vLLM, quantization, KV-cache sizing)
• Production monitoring, observability, and incident response
• Security compliance frameworks (NIST SP 800-171/53, CMMC, FIPS 140-3, RMF/eMASS)