Senior Capacity Planner

DigitalOcean•Seattle, WA

1d•Hybrid

About The Position

DigitalOcean is seeking a Senior Capacity Planner passionate about optimizing physical infrastructure for their Serverless Instance offering. This role is crucial for growing the Serverless Instance offering through physical infrastructure optimization. The Senior Capacity Planner will join a dynamic team focused on cloud computing and AI, reporting to the Capacity Planning Lead. The primary responsibilities include predicting customer demand shifts, optimizing utilization for Serverless Instances, and establishing processes and automation. The role also involves deploying new GPU hardware and collaborating with cross-functional teams to enhance physical infrastructure. The Senior Capacity Planner will analyze usage statistics for serverless inference models, identify optimization opportunities in server assignment, and make recommendations for GPU server allocation to maximize revenue and utilization. Building tools and processes for automated model-to-server assignment is a key objective. The role requires maintaining up-to-date knowledge of the GPU hardware market, tracking rack and power capacity globally, and partnering with DCOPS, Network Engineering, Product, and Delivery teams for GPU server deployments. Forecasting future GPU demand and identifying opportunities for hardware redistribution are also critical functions. The position involves creating dashboards and reports to monitor GPU utilization, spot capacity trends, and deployment status, as well as escalating any capacity risks or bottlenecks.

Requirements

5+ years of experience in capacity planning, infrastructure operations, or a related technical/analytical role
Strong understanding of the current GPU hardware market — models, vendors, availability, and performance characteristics
Familiarity with serverless inference workloads and spot capacity dynamics
Experience analyzing usage and demand data to drive resource allocation or optimization decisions
Strong foundation with SQL and data visualization tools (e.g. Looker, Grafana) to build reporting and models
Experience with, and strong interest in building, automation/tooling/AI agents to reduce manual analysis
Working knowledge of data center rack and power capacity planning concepts
Strong cross-functional communication skills, with experience partnering across DCOPS, Network Engineering, Product, and Delivery teams

Nice To Haves

Experience with data center inventory tools such as dcTrack or Netbox

Responsibilities

Analyze usage statistics across serverless inference models and spot demand to identify optimization opportunities in server assignment
Make and implement recommendations on GPU server allocation across models to maximize revenue and utilization
Build tools and processes to automate and streamline model-to-server assignment decisions, reducing manual analysis over time
Maintain a deep, current understanding of the GPU hardware market — new models, availability, price/performance tradeoffs — to inform procurement and deployment decisions
Track rack and power capacity for GPU infrastructure across global data centers
Partner with DCOPS, Network Engineering, Product, and Delivery teams to plan and execute physical GPU server deployments
Forecast future GPU demand (both inference spot capacity and physical infrastructure) to inform procurement timelines
Identify opportunities to redistribute or optimize existing GPU hardware placement across racks and cabinets
Build dashboards and reporting to track GPU utilization, spot capacity trends, and deployment status
Escalate capacity risks or bottlenecks before they impact GPU availability or revenue