Technical Product Manager - Soperator
Required skills
3–5+ years in Product Management, ML infrastructure/MLOps, distributed systems, or cloud platform engineering. Strong technical depth in distributed systems, cloud infrastructure, or ML platforms. Hands-on familiarity with large-scale ML training and orchestration tools (e.g., Slurm, Kubernetes, Ray ). Track record of shipping technically complex products with multiple engineering teams. Strong communication and stakeholder management across engineering, research, and customers. Experience with product analytics, data-informed prioritization, and experimentation. High ownership, high learning velocity, and comfort operating in fast-moving AI infrastructure environments.
Preferred skills
Experience with GPU platforms and HPC primitives: InfiniBand/RDMA, topology-aware scheduling , high-throughput storage. Practical understanding of modern ML training stacks: PyTorch, DeepSpeed, FSDP/ZeRO, NCCL . Familiarity with efficiency and reliability metrics: Goodput, MFU, failure modes, preemption handling, health checks . Exposure to large-scale LLM training/inference systems. Experience in observability, performance tuning, or SRE/reliability engineering. Customer-facing technical experience (solutioning, support, architecture advisory).
Technologies
Slurm, Kubernetes, Ray, PyTorch, DeepSpeed, FSDP/ZeRO, NCCL, InfiniBand/RDMA, topology-aware scheduling, high-throughput storage.
Responsibilities
Own the full user journey across Soperator clusters: Slurm workflows, dashboards, alerts/notifications, node lifecycle, and training/inference capacity management. Define product direction end-to-end: problem discovery → solution design → delivery → adoption . Lead deep customer discovery through interviews, usage analytics, and workload analysis to uncover high-impact opportunities. Drive execution across platform teams: compute, networking, storage, observability, IAM and etc. Translate frontier ML and infrastructure ideas into practical product capabilities for real-world GPU clusters. Define success metrics, prioritize roadmap decisions with data, and ensure measurable customer/business impact. Lead the open-source strategy and execution for Soperator: shape public roadmap themes, prioritize OSS-facing capabilities, and ensure strong adoption in the community.
Seniority
Not specified.
Domain
AI infrastructure, cloud computing, ML platforms, distributed systems.