Infrastructure Engineer — Seed-Stage AI Lab

Infrastructure Engineer

Build the cloud. Own the stack.

Confidential AI Lab
· San Francisco, CA

San Francisco, CA (On-Site) Full-Time
· In-Person Daily $185,000 – $235,000 + 0.25–0.5% Equity Seed-Stage
· 6-Person Team Visa Sponsorship Available

A confidential, seed-stage AI lab building reinforcement learning (RL) environments specifically for computer use — enabling AI models to train from their own experience rather than relying on human-labeled data. The founding team is six people strong, from top-tier universities, and based full-time in San Francisco.

The company raised a seed round under a year ago and is scaling fast. The current infrastructure team cannot keep pace with customer demand — this hire is a direct response to that growth pressure.

"The cloud architecture, networking, and database schemas you build directly determine how fast the product scales and how many customers they can serve. This is greenfield technical ownership with real autonomy — not a maintenance role."

You will own cloud infrastructure end-to-end — building it from scratch and reporting directly to the CTO. No inherited tech debt, no siloed teams. Just a small, highly technical founding group trusting you to architect the systems that power RL environments at scale for agentic AI.

AI capabilities are evolving on a 3‑month cadence. Your job is not just to service today's demand — it's to future‑proof the platform so it doesn't become a bottleneck as the models advance.

• Design and build production cloud infrastructure from the ground up — including networking, IAM, secrets management, and container orchestration
• Take ownership of Terraform modules and IaC workflows end-to-end, reasoning carefully about blast radius and failure modes
• Work closely with the founding team (reporting directly to the CTO) to architect infrastructure supporting rapidly scaling RL environments for agentic AI
• Build and maintain CI/CD pipelines and release processes that ship software reliably and stop broken changes before production
• Stand up and manage PostgreSQL databases from scratch — schema design, migration strategy, and production-scale operations
• Design and implement observability stacks (logs, metrics, traces) so the team can diagnose and resolve incidents quickly
• Future‑proof platform infrastructure as AI capabilities evolve, building beyond just today’s demand

• 5–12 years of experience operating production cloud infrastructure end-to-end
• Built production cloud infrastructure from scratch (GCP or AWS)
• Managed Kubernetes or container orchestration in production
• Designed CI/CD pipelines and PostgreSQL schemas from scratch
• CS degree from a top-30 program
• Strong Terraform or equivalent IaC at significant scale
• GCP or AWS production cloud infrastructure proficiency
• Experience at an early‑stage or high‑growth startup

• Deep GCP experience: GKE, Cloud SQL, Artifact Registry
• AI/ML-adjacent infrastructure environments
• Observability tooling:
  Prometheus, Datadog, Open Telemetry

• Engineering managers removed from hands‑on coding — this role requires daily deep technical ownership

Six‑person team, all technical, from Princeton, Harvard, and Dartmouth. You'll collaborate daily with the CTO and founders — no layers of management between you and decisions that matter.

The cloud architecture, networking, and database schemas you build directly determine how fast the product scales. This is not a support role — it's foundational.

From GCP architecture to Terraform modules to CI/CD pipelines — you'll make foundational choices that persist. Design docs matter here as much as the code itself.

The team skews young, highly technical, and high‑agency. They ship fast, learn new stacks in weeks, and genuinely enjoy working together.