Senior Product Manager - Compute Platform & GPUs

Every day, tens of millions of people come to Roblox to explore, create, play, and connect across 3D immersive digital experiences. AI models reshaping how the community creates, plays, and connects all run directly on Roblox’s Compute Platform.

As the Senior Product Manager for AI Compute Infrastructure, you will set the multi-year strategy and roadmap for Roblox's next-generation AI infrastructure—a rapidly growing fleet of GPUs and custom AI accelerators spanning core on-premise data centers, edge deployments, and public clouds. You will own the products and abstractions that transform raw hardware into production-ready AI compute: driver and firmware management, topology-aware scheduling, fleet-wide health, performance optimization, and custom Kubernetes primitives for frontier model training and inference.

Key Responsibilities

• Platform Vision & Roadmap: Own the strategy and execution roadmap for Roblox's Compute Platform, spanning Managed Kubernetes (Roblox Kubernetes Service), Managed Compute Services, and Fleet APIs across hybrid cloud environments.
• GPU & Accelerator Scaling: Build and scale distributed GPU/CPU infrastructure supporting both high-throughput LLM/frontier model training and ultra-low latency inference at global scale.
• Kubernetes & Control Plane Evolution: Productize custom Kubernetes Operators, Controllers, and Custom Resource Definitions (CRDs) to overcome core control plane bottlenecks (API Server, Etcd scaling, Kubelet behavior).
• Topology-Aware Scheduling & Reliability: Define placement strategies, hardware affinity constraints, preemption mechanisms, and resource health checks to eliminate wasted compute cycles and improve MTTR/MTTD.
• Cross-Platform Enablement: Partner closely with seven core platform teams (AI Research, Storage, Analytics, Data Engineering) to deliver foundational compute primitives that empower product teams.
• Cost-to-Serve Optimization: Balance hardware utilization, latency, financial cost, and feature delivery speed while maintaining extreme infrastructure reliability.

Requirements

Product Leadership & Technical Infrastructure Mastery

• 7+ years of professional product management experience focused on compute infrastructure, GPU platforms, or large-scale distributed systems.
• Deep practical understanding of Kubernetes control plane internals (API Server bottlenecks, Etcd limitations, Kubelet behavior) and experience productizing custom Operators, Controllers, and CRDs.
• Deep familiarity with GPU/accelerator architectures and hardware scheduling challenges (topology-aware placement, preemption, resource readiness validation).
• Practical knowledge of cloud-native networking, Container Network Interfaces (CNIs), and enterprise service mesh architectures.
• Demonstrated experience building compute platforms on cloud providers (AWS, GCP, or Azure) and hybrid on-premise data centers.

Soft Skills & Mindset

• Strong cross-functional leadership skills to bridge gaps between AI researchers, platform engineers, product teams, finance, and executive leadership.
• A builder mindset focused on rapid prototyping, continuous iteration, and applying AI to streamline infrastructure management.

Preferred Qualifications

• Direct hands-on background in AI model development, large-scale training, or distributed inference workflows.
• Experience with kernel-level driver development or custom hardware accelerator drivers.
• Experience designing or deploying autonomous agentic systems for compute and infrastructure operations.

Total Rewards & Workplace Benefits

• Compensation: Base salary of $280,540 – $330,950 USD + comprehensive long-term equity grant package.
• Healthcare & Wellness: Comprehensive medical, dental, and vision coverage options; mental health resources and wellness stipends.
• Financial Security: 401(k) program with company matching; life and disability insurance.
• Time Off & Work Culture: Generous PTO policies, observed holidays, and hybrid workplace flexibility.