AI Infrastructure Engineer

AI Infrastructure Engineer

Own and build the end-to-end infrastructure that powers Utilidata’s AI/ML models across edge, cloud, and data center deployments.

Responsibilities

• Lead the design and build of Utilidata’s AI inference platform, including architecture patterns, deployment standards, and operational practices.
• Own end-to-end model serving infrastructure for both on-prem and datacenter environments.
• Build fault-tolerant, high-performance inference systems at scale with emphasis on low latency, reliability, and cost efficiency.
• Collaborate with algorithms engineers to integrate power/inference data and configuration with power optimization algorithms.
• Optimize GPU utilization and inference performance across the hardware fleet (including NVIDIA accelerators).
• Establish MLOps best practices (CI/CD pipelines for model deployment, monitoring, and rollback across environments).
• Contribute to infrastructure roadmap decisions (build vs. buy, tooling selection, platform evolution).

Minimum Qualifications

• 5+ years of software engineering with a strong focus on AI infrastructure, backend systems, or distributed systems.
• Hands-on experience with AI model serving frameworks such as vLLM, SGLang, Triton, TensorRT, or TorchServe (or similar).
• Container orchestration/cluster management experience (Kubernetes, Docker).
• Experience deploying and operating infrastructure across both datacenter and on-prem environments.
• Strong understanding of GPU workloads and inference vs. training tradeoffs.
• Proficiency in Python; C++, CUDA, Go, or Rust a plus.
• Strong communication skills and ability to work cross-functionally in a lean environment.
• Willingness to travel up to 10% of the time.

Nice to Have

• Dynamo experience.
• Edge AI deployments or constrained compute environments.
• Infrastructure as code (Terraform, Helm).
• Observability tooling (Datadog, Prometheus, Grafana).
• Background in energy, utilities, or industrial IoT.

Compensation & Location

• $170,000–$210,000 base + stock options (commensurate with experience).
• Fully remote within the United States; periodic travel for retreats and key on-site engagements.