Staff Product Manager - Observability
Core
Own the observability product for Lambda's AI cloud, enabling customers to debug distributed training across 64 to 1,024+ GPUs by providing transparency on GPU health, cluster utilization, and InfiniBand fabric metrics.
Role type
Staff Product Manager (Technical Infrastructure/Platform)
Builds
GPU and cluster health dashboards, job-level telemetry, InfiniBand fabric metrics, and the internal platform telemetry layer for fleet operations.
Domain
AI Cloud Infrastructure / High Performance Computing (HPC)
Required skills
Product strategy, cross-team influence, data-driven decision making, distributed systems knowledge, SLI/SLO/SLA definition, stakeholder management, roadmap execution, technical writing, metrics analysis
Preferred skills
Observability product experience (Datadog, Grafana, Prometheus), HPC/distributed systems telemetry, developer tools/API-first products, PyTorch/NCCL experience, GPU fleet tooling (DCGM), usage-based cloud infrastructure
Technologies
InfiniBand, NCCL, PyTorch, DCGM, Datadog, Grafana, Prometheus
Responsibilities
Define and drive the observability product roadmap from single GPU instances to 1-Click Clusters; Decide which telemetry signals to expose in the console and API; Build diagnostic experiences to distinguish code/config issues from platform failures; Partner with SRE and fleet engineering to build the internal telemetry layer; Work with customers to understand debugging needs for distributed training; Align engineers and executives on the roadmap through evidence; Define success metrics and iterate on features.
Seniority
Staff, hands-on IC with strategic scope
