Finance Manager, Infrastructure & Capacity

Runpod

2d•$140,000 - $180,000•Remote

About The Position

Runpod is the AI Developer Cloud, serving over a million developers by providing a platform for AI experimentation, training, fine-tuning, deployment, and scaling. Having processed over 20 billion inference requests and recently securing a $100M Series A, Runpod is at a pivotal moment in AI infrastructure growth. The company operates as a remote-first team, emphasizing ownership, speed, and impactful work. The Finance team is responsible for building the financial infrastructure to support this rapid scaling. This specific role is crucial for managing the financial aspects of capacity and infrastructure, with GPU supply being the largest cost and a primary revenue constraint. The Finance Manager will oversee analytics related to capacity, utilization, procurement, and future availability, working closely with Infrastructure, Data Center Operations, and Sales. This is a foundational role, reporting directly to the Director of Finance, in a fast-paced, remote environment where precision and ownership are paramount.

Requirements

4+ years of experience in finance, FP&A, corporate development, or a related analytical role, with a strong foundation in financial modeling and forecasting.
Advanced proficiency in Excel/Google Sheets modeling, including capex, cash flow, and return analysis (NPV, IRR, payback).
Comfort working with large operational datasets and BI tools to turn raw usage data into decision-ready analysis.
Comfort driving ambiguous initiatives to completion with minimal supervision.
Strong negotiation and communication skills, with the ability to translate data into clear recommendations for business stakeholders.
Successful completion of a background check.

Nice To Haves

Experience in a high-growth startup or venture-backed technology company.
Experience underwriting capital-intensive investments — GPU fleets, data center capacity, hardware, or other asset-heavy commitments.
Experience with lease-versus-buy analysis, equipment financing, or revenue-share structures.
Experience in capacity, supply, or fleet planning, including utilization and fill modeling.
SQL or hands-on experience querying large operational datasets.
Working knowledge of US GAAP, including lease accounting (ASC 842) and fixed assets.
Experience in the AI infrastructure, cloud, or data center industry.

Responsibilities

Maintain a live view of available capacity across GPU types and datacenters, and serve as the company's source of truth for what we have, what's sold, and what's coming.
Drive fill on upcoming tranches of capacity, partnering with Sales and Infrastructure to line up committed demand ahead of bring-up.
Monitor utilization across the existing fleet, flagging underperforming SKUs, datacenters, and node groups and driving the pricing or placement changes that fix them.
Evaluate alternate methods of GPU ownership — revenue share, leasing, and direct purchase — and recommend the right structure for each opportunity.
Forecast cash requirements for new capacity commitments and existing lease renewals, and own the capacity capex plan.
Model the revenue impact of bring-up timing, tracking schedule slippage against plan and quantifying the cost of delay.
Evaluate and monitor non-GPU infrastructure investments, including storage and CPU capacity.
Build the unit economics of the fleet — contribution margin by GPU type, datacenter, and host — and set the pricing floors and return hurdles that capacity decisions are measured against.
Support commercial negotiations with hosts and datacenter partners, modeling term sheets, revenue-share splits, and lease economics.
Run scenario and risk analysis on the supply book, including GPU price deflation, the contract expiration ladder, host and datacenter concentration, and residual value or redeployment of aging SKUs.
Own the capacity reporting layer — fill, utilization, and margin definitions and dashboards — and the post-commitment variance analysis that measures each tranche against its underwriting case.