MLOps&Network Infrastructure Architect

We're seeking a senior-level architect with strong expertise in MLOps, DevSecOps, deep learning pipelines, and high-performance networking to design and maintain our GPUaaS platform's backbone. This role emphasizes low-latency, high-bandwidth networks for GPU clusters (e.g., RDMA over InfiniBand/Ethernet), cloud integrations, security, and automation. Ideal for someone with experience architecting ML systems in enterprise environments.

• Architect and deploy high-performance networking for GPU clusters, including RDMA fabrics, InfiniBand/Ethernet switches, VPC peering, low-latency interconnects, and traffic optimization for distributed AI workloads.
• Design and implement MLOps pipelines for deep learning: data ingestion, model training/distribution on GPUs, deployment (e.g., Kubernetes-based), monitoring, and CI/CD integration.
• Apply DevSecOps principles: embed security in infrastructure code, automate compliance, vulnerability scanning, and secure cloud/on-prem environments.
• Optimize AWS (or multi-cloud) setups for GPUaaS scalability, cost control, high availability, and fault tolerance in AI-centric systems.
• Diagnose and resolve network/performance issues in high-throughput GPU environments, including latency reduction and bandwidth maximization.
• Automate infrastructure using tools like Puppet for configuration management, Terraform for IaC, and orchestration platforms.
• Collaborate with cross-functional teams on integrations with data platforms, BI tools, and monitoring solutions (e.g., Prometheus/Grafana).

​