Infra Architect - Data Platform & ML Pipelines

General Intuition & Medal; New York (NY); On-site; USD 180,000 - 240,000; Full time

Job summary

General Intuition & Medal is seeking an engineer who can own orchestration and GPU clusters, from scheduling to I/O, and build preprocessing pipelines that turn raw gameplay data into training-ready assets. You'll work directly with the founding team, influence production systems across GI and Medal, and move between cluster scheduling, disk throughput, and inference latency in a single week.

This is a hands-on role for someone with substantial infra experience, who can own a system end-to-end

Qualifications

• Experience owning orchestration and GPU clusters, with responsibility for scheduling, utilization, and capacity.
• Ability to build preprocessing pipelines turning large raw data into training-ready data.
• Understanding of disk and network I/O as bottlenecks and how to optimize them.
• Experience optimizing inference in production, including batching and latency considerations.
• Comfort working across cloud providers with infrastructure as code and multi-region deployment.
• Proven track record of owning a system end-to-end and making architectural decisions that reach production.

Responsibilities

• Own orchestration and GPU cluster scheduling, ensuring efficient utilization.
• Build and maintain preprocessing pipelines for large gameplay data.
• Treat disk and network I/O as first-class constraints and optimize accordingly.
• Optimize production inference for latency and cost, managing serving runtimes.
• Work across cloud providers and manage multi-region deployments.
• Own a substantial system from design to production, with concrete, demonstrable outcomes.

Job description

General Intuition & Medal is seeking an engineer who can own orchestration and GPU clusters, from scheduling to I/O, and build preprocessing pipelines that turn raw gameplay data into training-ready assets. You'll work directly with the founding team, influence production systems across GI and Medal, and move between cluster scheduling, disk throughput, and inference latency in a single week.

This is a hands-on role for someone with substantial infra experience, who can own a system end-to-end